Yaesu FT-902-DM Mods
most of the mods applies also for FT-901

Bought FT-902 from LA6LCA in 1988, and it hasn't had any faults at all. Only problem is intermittent
PCB edge contact problems, but they are easily cured with windscreen wash liquid and vaselin. It is not
regarded as any problem that it has valves, they last very long - as provided that you are not a fool
who sit on the key and tune very carefully. I would have preferred sweep tubes as they last even
longer than 6146B, but spares have never been a problem, and the future was planned many years
ago. An OM told me that his 6HF5's in Galaxy V mkII had lasted 19 years, and still were 90% of
full rating - this is in spite that he was very active! Bought an FT-901 the other day for $125.-
as I needed a replacement for FT-7, but it was soon sold to a friend who had to replace
his FT-101ZE. I am told that the worst problem with FT901/902 is that the electrolytics
in the power supply may sometimes short, particularly those for higher voltages.

Block schematic for the IF section of FT-902 (FT-901).
This is the key to understanding signal levels and how the signals are routed
dependent on the different modes.

FT-902 (FT-901) IF passband information



IF frequency schematic showing carrier frequencies at the IF on the different modes, this is very important
to understand which frequencies are on RF compared to audio for RX and TX

FT-902 (FT-901) RX audio problem
The first mod I did was to improve the audio. It sounded awful, particularly bad with internal speaker,
I once had an FT-101B and knew it doesn't need to be so bad quality, FT-7 also works fb,
but it is easy to cure. Just change the C511 in PB-1705A to 47nF (.047). Fortunately it is conventional
component here, another Yaesu transceiver has 0.1µF instead of 0.01µF capacitor fitted - as a fault, but
it was SMD type, so it is two hours procedure to remove and replace.

FT-902 (FT-901) keying and the Key-click problem.
I suppose FT-901/902 isn't among the worst transceivers in this matter, but I'd
made up my mind to check the key circuit long befor I put it on the air for the first
time. Some serious construction (it is not called design!) faults were traced, with different
timeconstant for key-down and key-up



The ideal keying has same resistance for charge and decharge of the time-constant circuit.
I used this principle for 2m beacons LA3VHF and LA6VHF where the actual keyed unit
was the two last stages of a PYE exciter, but usually it is not so easy to implement a simple
circuit, but it is easier ways to do it.

3f) Improving keying for FT-901/FT-902. I always wish to have a crisp and clean cw signal, the received reports indicate that the result was good. It was some problems encountered, because the time constants for KEY-UP and DOWN would be quite different if only a capacitor was added, so the time constants needed to be divided into two parts with diodes to increase timeconstant when needed for a symmetrical keying shape. Suppose I have marked all changes with read colour, but it might be some more component values, they should be correct as they appear on the circuit diagram.
2) The key circuit isn't easy to understand, so I put a copy of it to the right, and at least, you may see the pin 16 connection at upper right (pin 14 goes to mode switch!).

PA secondary emission and flash-over protection.
It is also another problem with the screengrid circuit for 6146B's. The power supply is protected against flash-over from anode to screen grid, but secondary emission in the valve could cause damage to the valve. This may be allright to do in times when valves are easy to obtain, just go to any shop and pay some amount of money because you choose to protect the electrolytic capacitors instead of 6146B's. But now is 6146B's worse to find than electrolytic capacitors, so why not protect both?
The problem is a diode (D01) in series with L04 on the PA board - PB-1715 (No.17). A better idea is to connect a 10k resistor across the diode and a varistor from screengrid to ground og in the power supply, suggested type is SIOV14K150 which will limit the voltage to 240VDC.
I've never experienced any problems, so my circuit might work, or it wasn't needed, but I am not interested to prove it.



The power amplifier board PB-1715 (No.17), see more details on next circuit diagram


The handbook is not correct, it is a power reduction on 28MHz with a 2k2 resistor og the bandswitch. Here is shown my modification of the screen grid circuit.

FT-902 Speech Processor mods



Improving the FT902 speech processor. The original circuit had little effect, so I added 1N4148 diodes in anti-parallel.
The PROC. LEVEL also had little effect, this is greatly improved using a P-type JFET. 2N5462 will work, but J177 is much better.
On-the-air reports have been very convincing, it is no distinct clipping, and I am told that it is usually no problem to use the
RF-clipper (processor) in local SSB QSO's on 80, 40, 6, 2m, 70 and 23cm bands. It is approved by otherwise very critical friends,
like Morgan SM6ESG, although he don't like my old Microwave modules 70cm SSB transverter, hi.
The modification is described in more details in Technical Topics (G3VA), Radcom

FT-902 (FT-901) Reduced RIT-range (Receiver Iincremental Tuning)



The first and simple solution to reduce the range, particularly valuable for CW operation

FT-902 (FT-901) Mode-dependent RIT range



Usually it is practical to have different RIT range for CW and SSB, HF and VHF operation,
here is a circuit which suits my needs with 1.5kHz for CW and 5kHz on SSB, and most part
of my operation with this receiver has been CW and SSB on 144, 432, 1296MHz bands.

FT-901/902 with transverter:

RF output level is quite high with FT-200, FT-250, FT-101B, FT-901, FT-902. Suppose it is 100mW, while
most transverters hardly needs milliwatt level drive. Some transverters are constructed for high drive level, but
I don't see the point, so I've always modified it using a 1000W resistor in series with 2p2 capacitor,
instead of the original 10pF from PA-grid circuit. The larger capacitor may load the grid circuit and
tuning on higher bands may be upset. Also connected a 10K resistor to the transverter terminal, such
that +12V follows the signal during transmit, it simplifies the arrangement, no need for PTT lines,
and PTT lines to external equipment is very bad practice, if they are grounded the external
equipment may start the transmitter and cause problems.




FT-901/902 Transmitter modification to drive a transverter

"RX-antenna" modication.

Made external RX antenna connection for the FT-902 (FT-901), similar system to those I made for FT-250 and
FT-101B earlier. A cable from RX side of antenna relay goes to a BNC connector board, and another
cable goes back to the receiver input. So I may choose between which receiver I'll listen to, possibly
Drake 2-B or Drake R-4C, or use FT-902 with a transverter.



A simplified circuit shows what was done with FT250, FT101B, FT902 to provide RX-ANT
output and input so I can use which receiver I want with FT902 transmitter, and use FT902 RX
as a separate receiver for another purpose without any risk of transmitting into the RX circuit.
This is important feature for use with transverters, and often one receive on another band, than tx,
only 14 or 28MHz is common.

FT-902 (FT-901) transmit level adjustments.



Originally the transmit drive levels varies very much. FM has the highest drive level. Decided to equalize the
drive levels for the different modes, some minor modifications to different boards are necessary

A better application for the 'AMGC' switch.
The AMGC has little practical effect, it was mentioned by DL1BU when reviewing the FT901 transceiver in CQ-DL



Choose between SSB- and CW-filter in cw-mode

The AMGC switch comes to practical for selecting between SSB and CW filter, but few extra components are
needed to activate it. When the switch is operated the voltage in base circuit (see schematic above) is pulled
down from 12V towards +8V, and a 12V relay is energized. One might have used lot of time to look for
an 8v relay, but I had a small 12V relay, and it draws much less current than an 8V type. It is usually a good
practice to measure which voltage the relays need to operate, and which voltage it falls back on, so it is often
possible to use a 24V relay on 12V, but it must usually not have higher operate voltage than 9V for safe operation
on 12V system, and 6v for 9V operation.

Logarithmic S-meter for FT-901/902.

S-meter deflection is far from ideal, and some improvements were made with the original circuit, but
it was never somewhere near the desired result. I would accept that an S-unit varies between
3-7dB, which is not much accurate, so I looked for other ideas



It was a need to make more accurate measurements with the receiver. An S-meter where an S-unit could be
anything between 0.5-7dB is regarded as useless, but believed to be a feature for the average operator who
wish to give 59 reports to everybody. This circuit has 70dB range - provided it is set up correct, and this is
more than you need, I've never experienced that the meter peaked, but I've got some complaints
that I don't give a 59-report to an unreadable station.
It should be noted that the gain and current of the input stage must not be set too low, it is better
to attenuate the signal with resistors in the drain circuit! I suppose the reason is that the dynamic range
for the amplifier is reduced when it doesn't draw much current, so it will not drive sufficient when
signal to pass is "strong". The minimum voltage from MC3356 output is 0.6V. It is a bad idea to
use a simple diode here, the voltage drop with varying current may vary between 0.4-0.6V, you may
loose indicating range. It is much better to rely on a transistor B-E junction diode, or just use an
emitter follower. It also solves the problem with time constant in the detector, but this is an unsolved
problem, it could be better. It is also a problem with FT-902 that it is some false signals in the
IF and the logaritmic detector cannot be set to indicate signals lower than 2µV on the receiver
input connector. MC3356P is a very sensitive device. Another device used for similar purpose
is Philips TDA1576, but I suppose it won't work properly here because it is not sensitive enough,
or perhaps is it just what one need to operate the input stage with some decent current and gain?

PB1726A (FT902 VFO switch) obscure circuit diagram
Had a fault and needed to follow the VFO signal, but the handbook is obscure on this point.
Perhaps the worst is that details shown in the handbook are incorrect, so it was only one
thing to do; loosen the board and inspect the pcb to draw my own circuit diagram

FT-902 overall installation

PTT facility on the paddle for CW operation.

Have mounted a PTT-switch on the paddles for my shack. It simplifies operation and you have 100% control, a
nd easy to switch between RX and TX, and back again if you wish to check that you are not transmitting when you shouldn't.
Since it will always be a time delay when you throw the button to you can start transmitting it is no need for time delay, but
it may be another idea if you use no-control cw meteorscatter operation

The paddle is good quality, but haven't the faintest idea who made it, bought it from a very drunk Finnish OPR at
Aannaboda VHF meeting, who may be the producer, and definitely wanted to buy it back the next day, but I refused.

The Vibro-keyer paddle.
Since I worked at Rogaland radio in 69-70 we had some ideas what was the best paddles and it was two different makes to choose
between, Autronica and Vibrokeyer paddles. Bought my own vibro-paddle, and used it for a year, until I got a better German construction
which also could operate SQUEEZE, and it was used until 1987, when I got the Finnish paddle mentioned above. Vibrokeyer is a part of the
older vibroplexer. It seems possible to drill some holes in the iron block, and mount a PTT switch on the paddle, but I keep it only as a reserve.

Operation control units
The units mentioned above for operation on HF, 6m, 144, 432, 1296MHz. IF for 6m is 14MHz, while it is 28MHz for the higher bands.
For 13, 6, 3cm an IC202 is used to drive.

Es stand in der CQ-DL, Band 2, Testberichte
A really very interesting book from DARC showing testreports by DL1BU for the following equipment:
Atlas 210/215, CRF320, Drake R-4C, Dressler D200, FT-221, FT-301, FT-901,
IC-202S, IC-211, IC245E, IC-280E, IC402, IC-701, KDK FM 2015R, KDK FM2016E,
LT470 (Braun), Multi 2700, REIS SE200XL-A, SB-104, SE301 (Braun), TS-520, TS-820,

Vergleichtest FR101, FT-220, FT221, HG70D, IC201, Multi2000, Multi2700

Die Testberichte in Band 2 finden ganz großes Interesse bei den Mitgliedern. Der Band kostet einzeln DM 9,-.
Beide Bände können für DM 14,- bezogen werden.

Skandinavisk språklige artikler (Scandinavian language articles):

FT-901-DM Modifikasjon (2) Noiseblanker .......LA3ZL AR 79-08-212
FT-901/902 Forbedringer (2) ..............................LA8AK AR 90-07-206
FT-901-DM Modifikasjon (1) Klikk og NB ........LA3ZL AR 79-06-175
FT-901/902 Nøkling........................................... LA8AK AR 89-06-165
FT-901/902 Forbedringer (1).............................. LA8AK AR 89-02-038
FT-901/902 Noise-blanker ..................................LA8AK AR 89-02-039
FT-901/902 Processor (RF-clipper)..................... LA8AK AR 89-02-039
FT-901/902 forbedringer for HF/VHF.................. LA8AK OZ 93-07-407
FT-901/902 Forbedringer af - ..............................LA8AK OZ 93-07-407
FT-901/902 rettelse for....................................... OZ 7/93 OZ 93-11-661
FT-902 Xtal kalibrator kan gi intermodulasjon .......LA8AK AR 89-04-105

2005.03.05