This is the 40watt 400-470MHz linear to follow the first amp.
With the DTX1 output above 25% the amp red tx light comes on but there is no output. Adjusting the bias pot brings up the gain and it can be set for 5watts or more output.
These two amps when connected in series therefore will give 5 or more watts out on 70cms using 13.8volts.
Now for the bad news! I have only fully tested the cleaness of the output of the first amp on a mates spectrum analyser. The adjustment of the bias was done once I got home. As far as I can tell using my SDR etc. the system looks good but I will have to wait till I put it on the test rig again before I am certain.
The first amp may be good for lower frequencies too.
These amps are around on Ebay and if you buy them then don't forget the connecting wires at the same time as you will be waiting awhile.
Also a filter or two could be useful.
Further to my last post I have been doing some more experimenting...
The 'cheap' Chinese linear from Ebay (Ham Radio Power Amplifier for Interphone DMR DPMR P25 C4FM 40W UHF 400-470MHZ) can be modified easily to stay in the transmit mode. The bias can easily be increased to 5.5 amps current into the amp. Then it began hooting. This was easily cured (once everything else had been tried!) by connecting the output of the block directly to the output coax. The way to turn on the transmit mode was to put a 1k resistor between +ve supply and the diode detector output. The output of the RA30H4047M does not contact the PCB but goes straight to the coax and an earth wire is the straight lead near it.
The 1k resistor goes to the diode end of R5
Once this was done the lid was put back on and with the DTX1 directly connected to this amp, and set to 100% output, there was 5 watts of rf. The amp runs hot but ran for 15 minutes without a problem although the output dropped to 3 watts.
The spectrum is from a little aerial connected to my SDR and picking up the bleed from the dummy load. This is only an indication and proper testing is needed when I get around to G4GPD's house!
With the little amplifier in series between the DTX1 and the linear the DTX1 only needs to be on 2% output so really is not needed in this case.
I have noticed that the linear can hoot at switch on under some situations. This seems to be stopped by putting a large capacitor on the bias input to the RA30H4047M and making it come on slowly.
Hope this helps
Martin G8LCE IO70LD www.g8lce.com
I have made a couple of Linear amps using the blocks soldered to PCBs and had problems with the grounding and as you say they usually earth the block via the screws and solder tags to the PCB. I found that the output wire from the block going via the PCB to the output coax also caused a problem with one of my built ones, hence I altered this one to go straight to the coax which stopped it hooting. The output does rely on the supply volts quite alot. I think the bias probably is not as stablized as other simpler designs. I found that if the bias is set at 5.5amps with a 13.8volt supply then when the volts dropped to 12,5 the bias current was 4 amps or so. If the bias is set and then the supply voltage is raised then the current shoots up as the amp hoots!
The main thing is that this cheapish amp seems a reasonable starting place for a linear which has the RA30H4047M block with PCB heatsink and case which is similar to what people might be going to make. I have got it working to some extent and now others could add ideas. I have learnt from others and now I have added my findings for others to work with.
I had a nasty experience with the first one of this type of amp I made with a cheap PCB and 60watt block. It was difficult to stabilize as you say due to ground planes not working. I eventually mounted it on a large heatsink and all seemed well. One day there was alot of arcing and it died! I think the PCB arced internally but I could not see it. I then bought a PCB kit from Mini-Kits and another 60watt block. That one is still working (once I connected the output of the block straight to the coax not via the PCB) and gives a better output than this one from China but is not in a box yet!
There is an advert for a 80-90watt version similar to these with two fans. It could be useful if someone wants to jump in and spend £120 approx! That must still be cheap for that sort of power (getting on for 10watts DATV maybe?)
I will be trying the amp in the real world sometime and will report back
Has anyone any thoughts on the output from the SDR? It looks like the side lobes are 40dB down and I cannot find any problems outside the band.
I had a discussion with a mate about the spectrum. I used vMix and used an extra hdmi monitor as the external display of vMix Then using HDMI to video adapter, fed the same feed (via a HDMI distribution amp) to my 23cms TX and through GB3NQ. I could then slide the SDR display across from the main monitor onto the HDMI one so he could see the spectrum. Good old amateur radio being used for experimenting, but vMix is another thread......
What is it using to regulate the bias? It needs a regulator. Often a 7805 or similar. If its just using resistors that might explain the problem with supply voltage. The -40 dB shoulders are verging on unacceptable for 70cm if you are operating from a good site you could be causing interference. At 5W hopefully not but I recall it rapidly gets better with small decreases in output. Otherwise the SDR output looks good. Anything out of band would be a problem so a nice tight between the SDR and the amplifier filter might be a good investment, but make sure the amplifier input sees nearly 50 ohms at all frequencies or its going to try to oscillate again. A low pass filter on the output will make sure the 2nd harmonic around 870MHz is not too high. It probably won't be at 5W but it might if you plan to use at higher power for FM or SSB.
All agreed. I see these on Ebay for under £80 which is not too bad if you consider the cost of buying all the bits and then making a box etc. I had a look at the side lobes with a radio and they seemed OK so I had a little outing and tried 70cms DATV from 2 miles away. Not far but 90 degrees of my beam. Checked on g8lce.com and rx 3 and it worked. Just a tempory setup!!!
I also tried a 1.3GHz setup and that also work until GB3NQ came on!!
The setup on 70cms DATV was a Portsdown with 7 inch screen feeding a DTX1 and that at 100% output feeding the 40watt linear going into the 10 element yagi.
The signal was checked by looking at RX 3 on www.g8lce.com using the laptop and a radio modem on 4G
PS I'll have a look at the bias circuit....
Don't forget the regulator capacitors, it looks like there are some on the board but its hard to tell from the photo. There must be some drop in the bias circuit preventing you from fully biasing it with 5V. The nominal value on the pin is 5V or less. At 5V the datasheet indicates you should see about 4A of standing current draw and gain should be around 40 dB. Its pretty sensitive to bias voltage. How you have it might be enough for this application and if so, it will make cooling easier. There is a 3dB PI attenuator on the input side I can see made up of 3 resistors you could bypass. If not, you can raise the voltage regulator voltage up a bit and use the POT to control the bias. If you only have a 7805 there is a way to raise its output but I won't go into that just yet.
I was suprised by the voltage I measured on the bias pin of the block. I had seen that 5 volts was the maximum but when I raised the current to around 5 amps by adjusting the bias pot the volts were 7.5 aprox!! I assumed the circuit was as the usual scheme with a voltage regulator. It appears that the bias probably is controlled by the input rf level! Well that is used to switch between rx and tx so I reckon they thought they would use the same level to control the bias.
I will try two 5 volt regulators together tomorrow to get 10volts and put that on the bias pot and adjust it for around 4 amps.
The linear does seem to work quite well given the lass up I have made. It would be nice if it only needed a 8volt regulator, the bias pot slightly lifted off the PCB and 1k resistor added to convert this to a useable amp for the DTX1. I have the filters for before and after but they can be bought already made too. I know more earthing and capacitors may be nice but I am still trying to keep the initial conversion very simple.
It has been mentioned that the pcb could be thrown out and the one from Mini-Kits used but I still hope to keep it down to two components and simple soldering.
I think the gain will be raised with the extra bias current and I like the input attenuator as it could help stability. I will bear it in mind though as it would be an easy mod.
More to follow......
Don't put more than 5V on the bias pin. That's the maximum. The absolute limit is 6V. Putting 7.5V on it would probably damage it. I am thinking this might be one of the older modules, see this data sheet http://f5ad.free.fr/Docs_ATV/RA30H4047M.pdf If it is, then the 1.8A seems about right. The design changed many years ago and they are different.
Also not sure what you are intending to do to get 10V from using two 5V regulators.