The KA7OEI FT-817 pages
FT-817 Radio Modifications

Front panel view of my FT-817
A Front-panel view of my FT-817. 
Important Notice - Please read this warning:

This document describes procedures that could result in voiding of the warranty of your radio.

It also describes procedures and modifications that, if not precisely and properly carried out, could result in a radio that does not work or is damaged.

Furthermore, while reasonable efforts have been made to assure the accuracy of this information, it is possible that there are some errors, or that your (or my) radio is of a slightly different version and thus, differences may exist.  Additionally, any legal opinions expressed - while believed to be accurate - are not guaranteed to be correct:  You are expected to take total responsibility for your own actions.

It is assumed that anyone following suggestions made on this page is already thoroughly familiar with the technologies and techniques involved and possesses the necessary skill and knowledge to make their own judgment as to the appropriateness and validity of the the information.

If you choose to do any of the procedures outlined, you do so at your own risk.  You are solely responsible for any damage, voiding of warranty, or other harm that may come about by following these procedures.  It is very strongly recommended that, if you maintain your own radio, you thoroughly familiarize yourself with the FT-817 service manual.  If you don't have one, get one!

For reference, my FT-817 is the standard USA version with the first 4 serial number characters being 1D21.

Index of Modifications:

Improving the FT-817 Noise Blanker:

Note:  Not all '817s are afflicted by this problem, so you will want to check your radio out before even thinking of trying it...

In the past, I have seen several people mentioning that they thought the noise blanker in their '817 wasn't working.
Noise blankers and their limitations

Like many other radios, the efficacy of the 817's noise blanker on an AM signal is somewhat limited.  Why?  It has to do with the very nature of AM signals. 

First of all, the noise blanker works by detecting brief signal spikes that are much stronger than the average signal level in a wide bandwidth:  When it detects the spike, it will mute the signal path until after the spike has passed. 

Another problem occurs with noise-blanking on an AM signal:  With SSB, the signal is only there when the other person is talking but with AM, a carrier is always there.  When the blanker mutes the signal path (sometimes on modulation peaks) it punches a hole in the carrier, sometimes causing popping or clicking:  On SSB, that "hole punching" isn't as easily noticed. 

What's this about the wide bandwidth then?  A noise spike, by its very nature, covers a wide range of frequencies - so the same "pop" you are hearing on 20 meters is likely audible on 30 and 15 as well.  This noise spike is also likely to last only a few tens of millionths of a second. When a brief pulse like that passes through an SSB filter it gets widened (look here to see what happens to AC line noise through narrow filters) and becomes far more offensive. 

The noise blanker detects noise pulses before the SSB or CW filter - prior to their being widened - and mutes them before they even get to the filter.  Because the noise blanker is in a wider bandwidth (usually 10-30 KHz or so) it can be "desensed" by strong, nearby signals:  You have probably noticed that your noise blanker will seem to go on and off when a stronger signal is just up the band from you - and that is why. 

With AM signals, the dead carrier is only 6 db (1/4th the strength) of the highest voice peaks.  Compare this to SSB where the signal (theoretically) disappears completely during silence.  With AM producing a constant carrier, the noise blanker has more difficulty detecting what is a spike and what is normal modulation. 

Noise blankers, in general, have another problem:  They will often "clip" audio peaks, mistaking them for noise.  Some noise blankers do better than others at distinguishing the two - and some have adjustable controls.  Most of the time, you can count on the noise blanker on any radio to degrade the quality of the received audio - but this reduction in quality may be far less objectionable than the noise you are blanking!

I had noticed that the noise blanker in my '817 seemed to work reasonably well on "single pop" types of noise (the "sharp" pop occurring at "relatively" infrequent intervals such as that from a spark plug) but not as well on "light dimmer" type noise.  I then noticed that it didn't work too well when signals and the noise were weak:  If I turned on the attenuator and IPO, it seemed to be much less effective.

I have done a bit of digging, and have found two problems that can contribute to this.

Problem #1 and its resolution:

I broke out the service manual and proceeded to do the "Noise Blanker Alignment" step.  This step (which is on page 17 in my copy of the manual) describes measuring a voltage at Q1074 and adjusting T1027 for minimum voltage.  When trying to do this it immediately became obvious that this would involve removing the slug from T1027.  What was worse was that when this was done, the noise blanker didn't work at all.

A quick look at the schematic as well as some probing around with an oscilloscope revealed that this instruction is wrong - one is to adjust for maximum voltage.  After a bit of study and experimentation I have come up with a revision of these alignment steps:

  1. Turn on the noise blanker, connect a voltmeter to Q1074's base and connect a signal generator to the receiver.  Tune in the signal generator (on any HF/6M/2m/70cm frequency - but NOT FM broadcast.)  Measuring the voltage, adjust the generator's output from no signal to at least -30 dbm and note the voltage range.  Set the signal generator so that the voltage is about midway between that with no signal and maximum signal.  Off-air HF signals should NOT be used for this step!
  2. Adjust T1027 for maximum voltage.  If it gets too close to the upper range noted in step 1, reduce signal from the generator and re-peak T1027.
This revised procedure should work well and it is probably easiest to do it in the AM mode because the wider bandwidth will make it easier to tune in the signal generator.  I haven't verified that this procedure works in the FM mode, but it probably will.

Compared to how T1027 was set from the factory, the original setting was not even close!  I noticed another problem with my '817, however:  I could not peak T1027. The maximum voltage was with the core of T1027 set to maximum inductance.

In looking at the schematic, I noted that T1027 has a built-in capacitor across its secondary.  As it turns out, this winding is also completely accessible externally and has a "Q-spoiling" resistor (R1246, which is located on the top of the board) across it.

I was able to put a 5 pf capacitor across the secondary of T1027 (this could be placed across R1246 directly, or it could be connected directly to the appropriate lead of T1027) and was barely able to tune through resonance (a 10 pf capacitor would have probably been better - but putting this capacitor across was tricky enough that I didn't want to try it again!)

The result?  My noise blanker now works even on very weak signals near the noise floor of the receiver.  How well does it work overall?  I must admit that I'm strongly biased:  I also own a Drake TR-7 with the later version of the NB-7 noise blanker - and I have yet to see any noise blanker on any other rig that even comes close.  Comparing it to other "typical" Kenwood, ICOM, and Yaesu noise blankers, it seems to work pretty well.  It won't reduce the powerline buzz by >10 db or take out the light dimmer on longwave/AM broadcast band by 40+ db like my TR-7's blanker will, but it is quite effective against "spike" type noise and it takes the "edge" off of powerline buzz..  It could still be better, though...

Having said all of that, I'll say that the difference in operation between having T1027 "maxed out" and having it be able to tune through resonance was very minor.  It is likely that the blanker will work just fine if the revised procedure is followed, and one keeps in mind that T1027 may have to be set to maximum inductance.

Problem #2 - and its resolution:

After having done the above, the noise blanker did work better - but after a month or so, I still wasn't satisfied with the way it operated (or not) so I did a bit more investigating.

There was one thing that the noise blanker did that bothered me:  On a "noisy" frequency, when it was first turned on, it seemed to work really well for just an instant (1/4 second or less) - and then stopped working as well.


Let's take a look at the schematic of the noise blanker:


I decided to look at TP1072 - which is "conveniently" located on the bottom of the board (I'm being sarcastic about the "convenient" part...) by soldering a small wire to it and re-installing the main board. When the noise blanker was turned on while tuned into an unmodulated test signal, an oscilloscope showed that the voltage at TP1072 started at 6 volts or so and suddenly nose-dived - indicating AGC action.  As you'd expect, when the signal was removed, the voltage at TP1072 went back up.


I then noticed that the brief "working" period of the noise blanker was exactly coincident with the AGC action at TP1072.  The problem? TOO MUCH AGC ACTION!

Let's see here - what controls that?  There's R1343 - which is part of the RF blocking/filtering.  There's R1345, which would be used to set the AGC gain.  "Ok," I thought, "What's R1345's value?"  On the schematic, it is marked with *** indicating a "value selected in production" (that's my guess - it's not mentioned anywhere...)  So, I went looking for R1345 - and found its location on the circuit board.

It was never installed!

So, I moved the wire from TP1072 to the base of Q1076 (where R1345 would be) and installed a 100k potentiometer between it an ground.  Viola - at between 39k and 68k, the noise blanker worked pretty well!

The value for your radio will likely depend on the specific components, but I suspect that something in the 47k range would be about right:  This is something for which a trimmer potentiometer or menu adjust should have been provided.

Finding the right value is tricky:  Too low, and the AGC won't work - and the noise blanker will clip objectionably on noise peaks.  Too high a value, and it won't work any better than when you started!  You'll probably have to strike a balance between good noise blanker operation and that which causes annoying artifacts.  It is for this reason that many rigs have a "noise blanker" threshold adjust on the front panel!

If the noise blanker still doesn't work for you on weaker signals, you might want to reduce the value of R1455 slightly.  As it is from the factory, it is 220 ohms.  You probably don't want to change it to anything lower than 47 ohms - and lowering it at all tends to reduce the IF gain slightly, as some of the signal is "siphoned" off by the noise blanker circuit.  In other words, if you change it, you may want to do the alignment steps to re-calibrate the "RXG" and S-meter settings in the receiver when you are done.  I paralleled another 220 ohms chip resistor across R1455, yielding an equivalent value of 110 ohms.
My '817 noise-blanker modifications in a nutshell:

- I added a 5 pf capacitor across T1027 and used the (above) revised alignment procedures to allow tuning through resonance, thus maximizing signal into the blanker. 
- R1345 was added to slightly decrease AGC action in the noise blanker.  I used a value of 47k, but your resistance may vary. 
- A 220 ohm resistor was paralleled across R1455 to slightly increase signal to the noise blanker. 

Please note that these values were derived through experimentation and while they work on my '817, your '817 may require slightly different values.  (Please read the cautions below.)

Was this worth all that trouble?

I'd say YES.  The noise blanker is now quite effective against many types of powerline noise and ignition/fuel system noise (i.e. spark plugs and injectors.)  While it doesn't work as well as the blanker in my old TR-7, it does work quite a bit better than the blanker in some other recent-vintage radios that I have used.  It adds only a slight "edge" to strong SSB signals (not enough to notice, really) and it causes a small amount of distortion on voice peaks when listening on strong AM - but if you are being bothered enough by noise to need the noise blanker, its artifacts are probably far less annoying than the noise itself.


I have probably voided my warrantee on the radio by doing this:  I'm not too worried, as I have always fixed my own radios - and if it hasn't died within the first 2 months (yet) then it probably won't die within the warrantee period, anyway.

Note:  THESE COMPONENTS ARE VERY SMALL!!!  If this problem affects YOUR radio (you will need test equipment to verify if it does) then performing this change requires very good soldering skill, at least SOME experience with surface-mount technology, AND the equipment to do this.  You can REALLY screw up your radio trying to do this with the wrong tools.  If you do mess up the board, I don't know if they will repair it, or if you'll get hit with a very large bill replacing the entire board.  It is not practical to do this modification unless you have a service manual - and no, I won't copy one for you!

IN OTHER WORDS:  Don't do this unless you have the equipment, skill, and the fortitude to do it.  And before you ask:  NO, I WILL NOT DO THE MODIFICATION FOR YOU!

From what I have been able to determine in my readings of comments made by other users, this "problem" doesn't affect all '817s.  It could be that only a run or two of them (the first few digits of my '817's SN are 1D21) and it could be that this alignment step (hopefully they didn't use the English version of the manual...) was simply missed in the production line on a few units.  It could also be that there are a few unexpected component variations that prevented T1027 from tuning through resonance.  Although I'm not aware of it, I would hope that Yaesu has noted this potential problem and taken steps to remedy it.

Using rechargeable cells in the FT-817's alkaline holder:

Before going any farther, I should caution that you should not do this!  (I don't.)  But, since I know that some of you will (after all, doesn't "HAM" stand for "Haven't Any Money" sometimes?) here are a few tips on how to do it in a less-dangerous manner.  If you do this, you do so at your own risk!

The FT-817 is capable of charging a NiCd or NiMH pack, but a circuit inhibits it from charging when the Alkaline cells are used.  Why would Yaesu do this?

It is a matter of CYA - Cover Your Asterisk - or something like that...

NiCd and NiMH cells, when shorted, can produce currents in excess of 20 amps - more than enough to burn wiring and destroy components on the FT-817's main circuit board.  Because there is no fusing intrinsic to the Alkaline battery holder, they (apparently) chose to make use of anything other than Alkalines unattractive.

What about Alkalines, then, aren't they just as dangerous when shorted?

The quick answer is No.  The actual answer is a bit more complicated.

In general, shorted Alkalines do not produce fireworks in the same way that NiCd and NiMH cells do - their short-circuit currents are much lower, and they tend to self-limit rather quickly.  More importantly, Alkaline cells are typically clad in a very tough insulated metal casing.

In contrast, NiCd and NiMH cells are typically protected only by a single, thin layer of heat-shrink tubing that is easy to chaff.  Normal handling of these cells could expose their cases and if any of these cases (except for the "bottom" cell in the string - the one "closest" to ground) touches the radio's chassis (when a cell is being installed, or if it gets scraped against a sharp edge near the battery compartment) the result could easily be a damaged radio.

Having said all of this, if you are hell-bent on doing it anyway, here are a few things that may be done.  Again, if you do this, you do so at your own risk!

Getting more power from the FT-817

It is possible - through software modification alone - to increase the RF output power from the FT-817.  Because these "software" modifications are covered on other web sites, I'll leave it to the reader to use a search engine to find them if he/she is interested.  Nevertheless, I would recommend against operating the '817 at a higher power:  That doubling of power (an extra 3db) rarely makes a huge difference in the quality of the communications - and the extra power drain and stress on the radio just isn't worth it.  (It is worth mentioning that one can usually get that extra 3 db simply by using a better antenna or by moving the antenna/radio around into a better location, anyway...)

Before discussing that - why is the '817 limited to 5 watts in the first place?  First of all, in the U.S., there are different requirements concerning spurious emissions and harmonics for transmitters operating at low power levels.  Second, the '817 is considered to be a QRP radio - and 5 watts is the de-facto maximum power for QRP operation.  Thirdly, 5 watts is a common power level for portable radios (such as HTs, etc.) and is a reasonable compromise between current consumption and battery life.

Having said all of that, let's take a brief look at the power amplifier section of the FT-817.  Note:  For a more complete description, go to the "About the RF Power Amplifier" page.

The active devices in the output stage of the '817 (the "finals") are a pair of 2SK2975's operated in push-pull.  These devices are rated for a maximum power dissipation of 10 watts each - so at first blush, it might be assumed that two of them together are capable of 20 watts.

That is not necessarily so.

Just because a device is rated for, say, 10 watts, doesn't mean that it is a good idea to operate it at that power level.  There are several things to consider when rating a device:

Perhaps the biggest factor on reliability of any electronic device is heat.  Keeping in mind that it is the semiconductor junction itself that is producing the heat (this heat is, hopefully, efficiently transferred to the transistor's package - and then to the heat sink...) the case or heat sink temperature may not tell the entire story.  In the case of the '817, there are two transistors, placed almost next to each other - sharing the same heat sink.

Certainly, for a short time, the heat sink in the '817 (which consists of the frame of the radio) can handle quite a bit a power - but if you keep this up, heat buildup occurs and everything in the radio gets hotter and hotter.  Obviously, one would not want to keep up the sort of operation that would "cook" the radio and all of its components - but it becomes clear that you could were you not careful.

As it turns out, if the radio is operated from, say, 13 volts, it will typically draw about 2 amps when transmitting at 5 watts.  This means that the radio has about 26 watts of heat to get rid of - and most of this heat (but not all) is produced by the final amplifier.  Typically, the entire final amplifier section will operate with an efficiency of approximately 25%- 33% - which includes the power consumed by not only the final section, but the driver and predriver.  This optimistically means that for 5 watts of output, one can expect at least 15 watts of power to be consumed - and 10 of that 15 watts total will be in the form of heat.

Bumping up the power to 10 watts implies that the FT-817's frame will now be asked to dissipate at least 10 more watts of heat - for a total of 30 watts of power consumption by the final amplifier (assuming that the efficiency of the amplifier remains constant, that is.)

Practically speaking, reasonably careful operation on SSB or CW is not likely to excessively heat the radio, owing to the relatively light duty cycle associated with those modes.  FM operation, on the other hand, is going to be a problem:  Unless one keeps track of how warm the radio is getting, it could be easy to forget and push the radio and its components to its limits.  If one has the radio in, say, a carrying case - then not only can it be difficult to keep track of how warm the radio is getting, but the heat cannot escape - being insulated by the case - and the radio will get even hotter.

There is another factor to consider, too:  Peak voltage.  If one always operates into a matched antenna, then the RF peak voltage on the final transistors is likely to stay comfortably within the transistor's ratings.  Typically, this peak voltage can be expected to reach to about twice the RF supply voltage at most during normal conditions.   Because the '817 uses 30 volt transistors, this isn't likely to be a problem.

What about a mismatch condition, then?  It is possible for power, reflected down the transmission line, to additively reinforce the voltage waveform on the final transistor.  This can easily cause the voltage to peak much higher than transistor's ratings.

Also - if the amplifier weren't shut down by its SWR protection circuitry, it is possible that it would try to produce 5 watts - but since that power may simple be reflected back into the amplifier, instead of the amplifier trying to get rid of 10 watts in the form of heat (from our example above) it will have to get rid of all 15 watts.

This brings up another point:  The SWR protection circuitry in the '817 (like most other radios) does not actually detect SWR per-se, but rather reflected power.  To properly calculate SWR, one needs to know (at the very least) what the forward and reflected power is at that instant.  Typically, a maximum threshold of reflected power is set - and if the reflected power exceeds that threshold, the protection circuitry kicks in and reduces power output accordingly.

This makes sense, when you think about it:  It isn't going to be the SWR itself that is going to damage the radio's power amplifier:  If you ran a 1000 watt amplifier at only 1 watt - and had a 10:1 SWR, that amount of reflected power isn't going to damage the amplifier (assuming that it doesn't oscillate or something...) but a 10:1 SWR at 1000 watts can cause all sorts of problem.

The problem arises when one attempts to rely entirely on the SWR protection circuit.

Because the circuit detects a given amount of reflected power coming back, the circuit is not activated until it sees excess reflected power.  When you first key up, you go from zero watts of output - to whatever output power you have set.  At that instant, if there is an excessive reflected power, the circuit hasn't quite detected that condition, so the amplifier is operating - for a brief moment - at full power.  A very short time later (typically, milliseconds) the excess reflected power is detected and the power is cut back.  Note again:  This is true of practically any transmitter with SWR protection circuitry - not just the '817!

During those few milliseconds, it is possible that, due to a poor SWR, the final can experience high voltage conditions.  Even that brief instant of full power into a bad match can result in cumulative damage to the final amplifier devices.  Persistent operation in this manner can eventually lead to premature failure of the RF amplifier.

Finally, it is worth mentioning Rubber Duck antennas:  Typically, these antennas are a compromise all of the way around.  Often, they are electrical 1/4 wave antennas - but considering that the radio does not offer a very good ground plane to work against (on 2 meters and below, especially) these antennas are typically tuned for what the manufacturer considers to be "typical" conditions.

If you have ever put an HT rubber duck antenna on an antenna analyzer, you'd notice that some antennas only work properly (e.g. low SWR) when placed "near" something - like the body.  This makes sense, as one usually has the radio (and antenna) near the face when using it.  Other antennas (like the FT-817's rubber duck antenna being used on 6 meters) only work best on certain parts of the band - and this depends on how the radio is held and what it is near.

In other words:  Don't count on a rubber duck antenna to have a good match under all conditions.  This fact should be remembered especially if one has "souped up" the radio to operate with more output power.

Operating the FT-817 outside the Amateur bands
Using the '817 on the new 5 MHz ("60 meter") frequencies

Will the '817 operate on the "new" 5 MHz frequencies? 

The answer is Yes - if the radio is modified.  The modifications to permit this are covered on numerous other sites (If you can find this site, I'm sure that you can find the others...)  and are not repeated here. 

There are a few caveats to operating on these frequencies, however: 

- The 5 MHz frequencies are channelized.  This is most unlike any other amateur HF allocation (except for the Alaska Emergency frequency at 5167.5 KHz.) 
- These channels are USB ONLY and centered on the following frequencies:  5332, 5348, 5368, 5373 and 5405 kHz. 
- Emissions are limited to 50 watts ERP (not likely to be a problem for the barefoot '817 user.) 
- Operation is to be on a non-interference basis. (You operate that way anyway, right?)

The first of these is the one most likely to cause problems - if "common" PSK31 operating procedures are any indication.  Because the frequencies spelled out are center frequencies, operating with your readout set to those frequencies listed would be illegal!Why?  These frequencies are actually the centers of narrow "slivers" and thus extend above and below the stated frequency.  Therefore, USB operation on, say, 5332 KHz exactly (which is a band from 5330.6 to 5333.4 KHz) would mean that a chunk of your signal (which extends up about 2.5 KHz from your readout frequency) would occupy up to 5334.5.  If the opening of 30 meters is to be taken as any sort of indication, the FCC will be ready and waiting to hand out tickets to those who weren't paying attention to the rules! 

Using the example given in the FCC Report and Order the actual frequencies shown on your readout would be 1.5 KHz lower than the aforementioned center frequencies.  These would be:  5330.5, 5346.5, 5366.5, 5371.5 and 5403.5 KHz, USB only!  You can find an updated color band chart on the ARRL web site.

Why did the FCC list just the center frequencies originally?  Well, I suppose that they assume that we hams can subtract! 

Disclaimer:  It is up to you to determine when and where you may operate legally on any new frequencies.  Don't hold me responsible for any typos or misinterpretations!

There are a number of variations of the FT-817 and not all of these versions can be modified in the same way and because of this, I'll leave it up to you, the reader, to find those modifications.  Since the web sites on which these modifications are found are constantly changing and being updated, it would be best to do a search and find the most up-to-date information for the version of radio that YOU have rather than try to rely on any incomplete, out-of-date information that I might have swiped from someone else's web site!

There are a few legitimate reasons why anyone would wish to operate the FT-817 outside the amateur bands, and some of the reasons why someone might do this include:

A few things that the FT-817 cannot be legally used for. Period.: (Note:  It should go without saying that if your operating on the above frequencies is done to save lives, then you aren't likely to get in any serious trouble...)

The above statements apply for those areas under U.S. Government jurisdiction.  The laws of your radio authority may be different.  Check your local listings...

So, where are the modifications?

Once again, I'll mention that such modifications are not available at this web site as there a a lot of people who have gone to the trouble of doing a good job maintaining their own up-to-date web site, covering the different flavors of FT-817 and the markets for which they were intended.  I'll leave it up to you to search for them than relying on possibly-outdated links that I might have put on this page.

There are no known modifications to allow the '817 to transmit in the 76-108 or 108-137 MHz range.  Furthermore, there are no known modifications that even allow reception above 154 MHz, below 420 MHz, or above 470 MHz.  (No, really!)

Using the FT-817 on Aeronautical frequencies

I get a surprising amount of email saying something along the lines of:  "I'm a licensed pilot, and I'd love to use my FT-817 as a backup rig.  How do I do that?"

First of all, doing so (i.e. transmitting) would be illegal - except (possibly) in the case of using it in a life-threatening emergency.

The answer?  You can't.

While the FT-817 may do a good job of receiving the Aeronautical band (some versions, anyway) it simply lacks the RF circuitry for transmitting on this band:  There are tracking filters that are present for transmitting on the 2 meter band, but are completely absent for the aero frequencies.  Furthermore, the software would appear to be well-written to prevent even the attempt to transmit in that frequency range.

What do I think about the '817?

Despite what has been said on this page, I really do like my '817 - warts and all.  I might point out that like any complicated piece of equipment, it is unreasonable for anyone to "get it all right" the very first time - and I'm not overly concerned.  Also, one must keep in mind that this is not the top-of-the-line radio.  Cramming thus much "stuff" into such a small volume is likely to require a few compromises.  No if someone doesn't learn from feedback and/or their mistakes and improve their future products, then I start to worry...

Work continues on this page - please revisit soon!

Yet another notice:  The information contained on this and related pages is believed to be accurate, but no guarantees are expressed or implied.  The information on this and related pages should be considered to be "as-is" and the user is completely responsible for the way this information is used.  If you find information that you believe to be incorrect, please report it via email.

The KA7OEI FT-817 "Front Page" - This is, well, the "front" page of the '817 pages here...

Any comments or questions?  Send an email!

This page maintained by Clint Turner, KA7OEI and was last updated on 20090824.  (Copyright 2001-2009 by Clint Turner)

Since 12/2010: