A Front-panel view of my FT-817.

About the Power Amplifier:

On of the most amazing things (to me, anyway) about the FT-817 is that the same power amplifier chain is used for all of the amateur bands from 160 meters through 70 centimeters.  This is made possible through the use of MOSFETs.  Unlike bipolar transistors, the gain of MOSFETs doesn't decrease at a 6db/octave rate as you go up in frequency.  What does make broadband amplifier design with MOSFETs more difficult is high gate capacitance:  At 160 meters the gate capacitance (45 pf per device for the devices used) is nearly invisible, but at 450 MHz, this same amount of capacitance complicates design.  In an extremely broadband design such as this, the gate capacitance amounts to a swamping reactance at the high end of the frequency range, reducing effective gain. Note:  Contrary to rumor, the power amplifier of the FT-817 is NOT class-A - more like Class AB-something..

The original FT-817 finals used a pair of 2SK2975 MOSFETs in push-pull configuration and the ratings of this transistor (which is designed specifically for VHF/UHF use) are as follows:

• Drain/Source voltage:  30 volts
• Gate/Source capacitance:  45pf (Vgs = 10 volts, Vds = 0 volts)
• Drain/Source capacitance:  80pf (Vgs = 10 volts, Vds = 0 volts)
• Power output:  8 watts (typical) at 9.6 volts, 1 watt input at 450 MHz
• Device dissipation:  10 watts (when soldered to a 14x2.8x2mm copper block)
• Power gain (at above conditions) is at least 8.4dB
• Power-added efficiency:  50-55%

Taking a look at the schematic, one can see that the power amplifier is optimized for VHF/UHF operation and has been "de-optimized" somewhat (in terms of gain) for  HF/6 meter operation.  Now, before one gets worried about this, one must realize that owing to the way the physics works, if you make a broadband RF amplifier with plenty of gain at VHF/UHF, it will have way too much gain at lower frequencies.  Because of this, degenerative feedback is employed to reduce the gain at HF and 6 meter frequencies to a "reasonable" level - otherwise the gain would go to astronomically high values (making for an unstable circuit) at the lower frequencies. 

How much power is the amplifier capable of?  It all depends on what you are looking for.  At 70cm, the RF amplifier chain as a whole isn't capable of too much more power than its ratings - 7 or 8 watts maximum at 15 volts is the most one could expect and have reasonable linearity on SSB.  Most likely this is due to a lack of excess gain in RF stages prior to the final amplifier section

At HF the story is different:  There is enough gain such that the PA is capable of being driven to well over 10 watts.  This doesn't mean that doing so is a good idea:  Linearity under all operating conditions may be suspect as the RF amplifier stages will happily put out this sort of power at 15 volts - and do so cleanly - but at 8 volts, trying to get this much power will likely result in terrible linearity (i.e. splattering) in SSB and AM.  One also needs to watch the dissipation of the radio's chassis at higher power levels:  Running higher power levels on SSB or CW (where the duty cycle is low) is probably just fine, but running FM at such power levels will likely result in overheating (shortening component life) - especially for long-winded people

"Does the FT-817 have thermal  protection?" The FT-817 doesn't have any thermal (overheat) protection per se:  Ideally the user would note that the radio is getting too hot during use. It isn't a good idea to run any radio at full power while obstructing its cooling:  Also, engaging in long-winded QSOs and using full power while the radio is enclosed in a carrying (restricting ventilation) case isn't a good idea.  One slight amount of implicit thermal protection is the fact that MOSFETs (which are used in the '817's power amplifier) have the property of their gain reducing with increasing temperature.  Theoretically, this means that the output power will drop as they get warmer.  In reality, this "thermal foldback" property will probably not prevent a negligent operator from overheating the radio and damaging it.  Many radios (the FT-817 included) have a feedback circuit used for controlling output power that will increase the drive power to the power amplifier to compensate for its drop in gain.  In a nutshell, don't obstruct the flow of heat if you are going to run your '817 (or any other radio!) for long periods at high power.  Also, the FT-817 has a Time Out Timer (menu item 49) that can be used to help prevent an accidental unintended keydown (i.e. sitting on the microphone) as well as minimizing long-duration transmissions.

Finally, it must be mentioned that the transmitter's harmonic and spurious emissions may not pass legal muster at power levels over 5 watts.  This has to do with the FCC limits on spurious radiation at power levels over 5 watts as well as possible increases of spurious signals (e.g. harmonics) as the power amplifier is driven harder.

The FT-817 and the "Blown Finals problem"

Almost from day one there has been a report indicating that the FT-817 has a tenancy to "blow finals."  In the years since the '817 came out, reports of this tenancy have persisted but attempts to get to the root of the problem have not yielded any singular cause of this reported problem.

Does the FT-817, in fact, have more of a tenancy than other rigs to "blow" its finals?  It's hard to tell, actually.  Before going further, it's worth pointing out several things that can contribute to this reputation - be it deserved or not...

• No-one really knows how many '817s have blown their finals.  For understandable reasons, the manufacturer has not made it known how many different units have suffered from this problem.

• Nowadays one can keep in contact with literally thousands of other users via the Internet:  You will certainly hear more reports of failures than "lack of failures" because, as everyone knows, bad news travels  more broadly than good news.

• Being that the '817 is a portable radio, it may be more likely that it is exposed to marginal antenna systems with poor matches.  As every operator should know, the "SWR Protection" circuit is designed to protect the user from the effects of the occasional "oops" rather persistent operating into a terrible mismatch.  Remember:  The SWR Protection circuit must "see" a bad load before it knows that it exists.  In the instant "before" it detects the bad load, the finals are exposed to mismatch conditions.

• For better or worse, the '817 is easy to "modify" to increase the output power.  It really cannot be argued that increased output power will not put more thermal and electrical stress on the devices and that it is likely that this added stress increases the probability of failure.  What is unknown is the degree that this increased stress affects the reliability of the amplifier.
  

Interestingly, Yaesu has changed the design of the power amplifier several times since the '817 was introduced.  As you might expect, it is normal for a manufacturer to make modifications to currently-produced products to improve them, to accommodate parts availability, and to make manufacturing more cost-effective.

As noted above, one of the early major changes of the '817's PA board was the replacement of the 2SK2975 (apparently being phased out by the manufacturer) with a RD07MVS1 - a slightly "hotter" device in terms of gain.  Because of its slightly different properties, it is required that several other components be changed in addition to just the transistors and because of this, it is usually more cost-effective to simply replace the board than try to modify an old one to use the new device.

Other changes made to the PA boards have to do with the change of the input transformers (after lot 09) and output transformers (after lot 74 - when the new finals were being used.)  I don't know the official reason as to why all of these were done, but reports indicate that it may have been done to improve overall stability of the amplifier as well as to minimize losses at the extreme high and low ends of  the frequency range of the '817 PA:  Remember that this thing operates over a 250:1 frequency range - nearly 8 octaves!

One more reported change has to do with D3052 (located on the underside of the board) which is used as a transient absorber and is connected across the RF path on the "wiper" of the "Front/Rear" antenna select relay.   The reported change is to substitute one that is much faster than the previous unit - something that would be done to protect the radio against high-speed transients on the antenna line, such discharge of wind static (something that can occur in a vehicle as well as on a home station,) a "zap" that might occur when one connects a coaxial cable, when one touches the radio or antenna with a "body charge" of static, or even a nearby lightning strike.

Whether or not these changes are a tacit acknowledgment of problems by Yaesu is unknown.  I would presume that such changes were done to improve the performance and reliability of the FT-817 - and it is always encouraging to see that a manufacturer does continues to improve on a product that is in current production.  As for my '817 (a rather early unit) I have yet to have any problems with the finals - which, for the vast majority of the users is the case.

Replacing the FT-817's power amplifier yourself:

In the somewhat unlikely event that you do have to replace the finals in your FT-817 (being that the vast majority of '817 users have had no problems at all with theirs) it is worth knowing that a hobbyist with reasonable soldering skills, modest equipment, and attention to detail can, in fact, do the job his/herself.  An example of this was Mike, PA7XG, who undertook the challenge - despite initial reluctance to do so.

Of the task, Mike said "Thanks for your homepage. It convinced me that I could do the job. (It's tricky work for a senior citizen (me) with bad eyes and dodgy fine motor functions but otherwise straightforward.  Nor would I recommend it for a rank beginner..."

Mike kindly emailed me a file that describes in detail how it can be done, along with permission to post it for all to see, and I do thank him for doing so:

• FT-817 Final Replacement - .PDF, 21k

Addendum:

More recently (March, 2012) I received information from another amateur who undertook the project of replacing the finals in his FT-817 and he noted a few differences from the above procedure no doubt reflecting the more recent revisions of the radio and PA board:

• In the above document, a black, plastic adhesive shroud was noted:  This doesn't seem to be present in all FT-817's.

• The above document also refers to unsoldering ground pad "B" and a copper spring:  Some versions of the final don't have either a connection to ground pad "B" or a copper spring.

• The above document mentions using a supply voltage of 11.5-12.0 volts for bias adjustments while the FT-817 service manual specifies 13.8 volts.

• When it comes to setting the bias, there seem to be a number of different specifications to be found:   Practically speaking, the amount of bias isn't terribly critical and its exact value will depend somewhat on the supply voltage and it will tend to inversely vary with the final transistors' temperature, so don't worry about it being 10 milliamps off the aforementioned specification.
• First, sett both VR5401 and VR5402 for minimum current on the PA.
• After this is done, VR5401 is then adjusted for an idle current of 45 +/-2 mA.
• THEN adjust VR5402 for an idle current of  85 +/-4mA - but don't readjust VR5401.
• Considering that at DC, both final transistors are in parallel, one would adjust the second control (VR5402) to approximately double the current. 

• As of May, 2013, reports indicate that the part number for the PA board is "CB 133001" and it cost $43.88 from Yaesu, U.S.A. not including shipping/handling which, for one reader, brought the price up to $52.31.  According to him, the swap out took less than two hours and, as other email attests, quite easy as long as one pays attention to detail.
• When replacing the final unit, make certain that there is voltage on the "13US" line (which may be labeled as connection "E") and is connected to J3003 - the 3-pin connector that feeds to the RF/PA board from the main board.  This should be an "always on" voltage and if it is not present, suspect a burned trace on the "main" board or elsewhere.

The analysis of the SWR indication/protection of the '817 has been moved to the "VSWR Metering" page.

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

Any comments or questions?  Send an email!