This mods you get more power on 4m from the FT847
After mods I have 60W output and current is now only 12-14 amps
Here is a copy of the original artikel from www.70Mhz.org
by Keith Naylor G4FUF
There is a lot of information on the internet about using the the FT-847 on 70MHz, unfortunately almost all of it is incorrect. This memo was originally produced in April 2000 for private circulation.
The modifications described here are valid for models with serial numbers
"9E" (May 1999) or later. Prior to that date the firmware is at an earlier issue
and has not been verified.
If you are unaware, YAESU date codes are the year of manufacture followed by the calender month as a letter code. As examples, FEB 1998 = 8B, MAY 1999 = 9E and JULY 2001 = 1G.
Note that serial numbers 9E and onward models have additional circuitry in the way of wire modifications added to modify the APC loop on TX. This is to reduce the output power to circa 15W between 54-76MHz. This was due to too many blown PA's.
Allowing the APC to be open can result in anything from 25W up to 40W output, albeit at >20A supply current.
First you must examine the boot condition for the CPU. The correct product
configuration codes for a 70MHz model are 0x2C or 0x6C(hex). This is set by 6
hardware jumpers adjacent to the CPU backup battery. The jumpers are numbered
1(LSB) to 6 (MSB). These settings are independent of the CPU reset function,
therefore a full reset is not required after modification.
For code 0x2C or 0x6C the settings are:
1 2 3 4 5 6
0 0 1 1 0 X
0=closed (a short or zero ohm resistor link)
X=open or closed depending on wideband 37-76 TX or just 70.00-70.50 TX
Please check and confirm the configuration. The picture shows the battery and the adjacent jumper location:-
However, to be sure of the firmware code one must also verify that the serial
boot EEPROM is correct. You will need to measure the voltage at several test
points on the RF-UNIT. The filter select data is serially clocked into I.C.
When 70MHz is selected and the radio is on TX verify the following (refer to circuit diagrams for all references):
TP3012 = <0.4V
If the jumper selection at the CPU is incorrect then the wrong BPF will be
selected. It may appear to work but very bad spurious outputs in the range
40-45MHz will be present when on 70MHz.
Models after 9E should be OK. Models prior to this require some additional switching added around Q3018 unless the firmware is reloaded.
I have investigated the power output issue further after requests from local stations to modify some radios. Further examination reveals an apparent design flaw in the PA which I had previously overlooked but found by accident. This now permits a non-invasive modification to substantially increase the output power and efficiency, whilst the radio may be restored to its original state at any time.
Inductor L5006 (see circuit) which is part of an m-derived low-pass filter, has a rogue spurious resonance close to 70MHz. This can be easily verified with a simple GDO.
The modification moves the resonance down slightly by inserting a small ferrite core into L5006. This inductor is a 5-turn coil approx 7mm in diameter adjacent to the TX/RX C/O relay RL5001. (to the left of the relays in PA photo).
The core I used was taken from an old TOKO S102 series VHF/UHF coil and is held in place by a small dab of non-corrosive bath sealant, or acrylic from a hot-melt glue gun.
The power output of the unmodified radios was 20W. After modification, the power output had increased to 65W/75W respectively.
Some FT847s have a wired mod fitted around the "50" VSWR-FORWARD pot on the RX-IF board. This wire connection to the "50" pot can now be removed, as the ALC will limit at the nominal 100W 50MHz level). The limiting factor now is drive power as measured out of the exciter (AF-CTRL) board. The original current consumption at 20W output was nearly 20A. With the modified L5006, at 75W out it is now just 17A (There is still significant power lost in the main PA filters which have not been touched - see section 3 below for a modification). The power output at 50MHz remains at 100W, with a small drop in efficiency estimated as ~0.35dB loss.
After 15 minutes' continuous transmission at 100W on 50MHz, the modified
L5006 core is just warm to the touch.
The second harmonic level at 50 and 70MHz is >55dB down on both radios tested, essentially the same as before the modification.
The output of the final stage is coupled into a LPF comprising C5057, C5058, C5061 and L5006. This filter has a loss of 0.9dB at 70.2MHz.
When operating on 70MHz the low-pass filter LPF is selected. LPF is a
7th-order Chebychev filter which has been tweaked to put two deep
attenuation poles at approximately 90MHz and 110MHz (to cover the EU broadcast
The component values as fitted on 70MHz models are:
C5123 33p fitted on at least 8G, 8J, 8K S/N models. Deleted on 9E onwards.
The filter characteristics are a compromise at 70MHz and are measured as:
50MHz: insertion loss 0.13dB; VSWR 1.12.
70MHz:(C5123 fitted) insertion loss 1.4dB; VSWR 1.2.
70MHz:(C5123 omitted) insertion loss 1.2dB; VSWR 1.4.
The total loss in the PA-UNIT output assembly at 70.2MHz is approximately 2.5dB (2.3dB) in the TX path and 1.4dB (1.2dB) in the RX path.
See also the supplementary information in the Equipment section of the Forum.
Between 54-76MHz, a broadband RF amplifier is selected. This has a
noise-figure of approximately 11-12dB on 70MHz.
As a comparison, the RX gain is approximately 8dB lower on 70MHz than 50MHz. Note that increasing the RX I.F. gain via the menu will not improve the S/N ratio of the RX. It might sound healthier but do not be fooled.
Most later models have a small piggy-back PCB mounted on the RF-UNIT which contains an attenuator and LPF assembly for use below 1.82MHz.
This piggy-back board may be replaced with a switched 70MHz pre-amp using a BF981 or similar device. Alternatively, the PCB can be modified into a 70MHz pre-amp and use the on-board relays but this is a one-way ticket, requiring many tracks to be cut.
A BF981 or similar MOSFET will provide a noise-figure of less than 0.5dB. With the LPF losses present in the PA module ahead this still represents an overall noise-figure of less than 2dB which is more than adequate at 70MHz.
Given a number of requests, Ian G4EZP and I are prepared to manufacture a batch of pre-amp PCBs. This PCB is a replacement for the Filter/Attenuator daughter-board which is fitted on the RX PCB of all European/USA imported models. If your radio is a 'genuine' 70MHz model, then all the switching voltages are already present (if not, a modification is available to add a single transistor and a wire mod).
If you are interested, please email
If you have problems with this email address, an alternative can be found on QRZ.com.
I intend to supply just a kit of parts, unless someone pleads for help!. The cost will be around £10 for the blank PCB alone: I will be able to give a firm price for a complete kit in a couple of weeks if there is sufficient interest.