Dutch RF - 6w DRFS06

[Shedbuilt]

Great news, and great work so far.
There is a pot for FET bias, and a pot for power / drive, which varies the voltage to the sga device. I can't look at the schematic either at the moment, because I'm not on a PC, but I think both use transistors. The "Power" certainly does, and also has an enable / disable input from the PLL (based on PLL lock detect). I take it the lock LED (and display indication) were on at the time you tried to measure. If so, take one at a time, and try what Pjeva suggested; turn the pot while checking the voltage. If you can't get any voltage, check that all the control transistors are in the right positions, and the right way round. There are a mix of NPN and PNP, so they have to be in the right places. Beyond that, it's going to need fault finding at component level (measuring DC voltages on the transistors. I'll try and download the schematic at home when I can, to try and help if needed). For the power control, you could also try disconnecting the end of the choke connected to the sga, and check voltage again - in case the device has gone short. It's likely that you have some blown semiconductors from the reversal, but one step at a time !!

EDIT: Can't quite remember if the power and bias use the same Zener. I think there is a 6V2 Zener for one or both. Might be worth checking it's the right way round, and the voltage across it too.....
I'll try and post the link to the documentation, when I'm on a PC too - where I found the schematic. There's a documentation link from the item on the website / shop.

[LAZY]

Ok, LED Lock light lit, and shows LOCK on LCD. The bias and power pots do nothing to change the the voltage of the SGA IN or Gate of RD06.
Disconnecting the choke = No change.
All the Transistors are in their correct locations and correctly orientated. How do I test to see if they are damaged? I see BC557 Q6 shows on the scematic as RF Off 0=off 1=on, could this knackered and causing the problem?
Not sure if it helps diagnose, but the RF IN of the SGA shows Negative Millivolts.
There are 2x Zenner Diodes 1x 6V2 Zener Diode 0,5w and 1x 5V1 Zener Diode 0,5w both are correctly installed. Not sure what the voltage of these should be.

I really appreciate your help and input guys. Even if I don't get this one working I think you both deserve an additional star ranking!

Lazy.

[Shedbuilt]

OK, 5V1 means 5.1V. This is the Zener for the logic supply, which is running the PIC, PLL etc, so this would seem to be ok. The 6.2V Zener is only used by the FET bias circut. It's not providing power to the sga. The power control circuitry provides the power to the sga, and also to the bias circuit for the FET. Q6 needs to be turned on, to start things off. When Q6 is turned on, it will provide power to the power control pot. The power control provides a varying voltage, via R17, to Q5. Q5 regulates the circuit output voltage, by controlling drive to Q4, which provides the output to the sga device, and to the bias circuit for the FET.
I would probably start by measuring the voltages on the pins of the power pot. If there is no voltage on any of the the power pot pins, then either Q6 is blown, or not receiving the on / off signal from the PLL chip. Check this too if you can....
If you're feeling brave, and confident of doing it without shorting pins / legs together, you could measure voltages on each leg of each of the three transistors, and post them up. Even better if you can do identify base, collector and emitter voltage for each.

[Shedbuilt]

OR, and again if you have voltage to the power pot, make sure that the voltage on both sides of R17 changes as you adjust the pot. Then check the voltage at the junction of R18, R19 and the Emitter of Q5. This should be about 0.7V below (more negative than) the voltage going into the base of Q5, from its connection with R17 (also connected to one end of R20).
If that's all good, see what's happening on the collector of Q5, which is connected to the base of Q4.
The negative millivolts seen on the input of the sga, with any luck, may mean that there's RF there, confusing your meter. If so, it's a good indication that the MSA device is working.....
In this post and the last post, where I mention voltages, I mean with respect to ground - with the -ve lead of your meter connected to ground......

[Maximus]

The 15W version from a couple of years ago. Fantastic piece of kit!

Edit: not sure why the pic went sidewards??

[teckniqs]

What's the maximum power you could get out of the 15w version? I heard people have had 20-25w on some frequencies using those boards.

[pjeva]

What's the maximum power you could get out of the 15w version? I heard people have had 20-25w on some frequencies using those boards.

I made one of those last year, when you properly filter harmonics, I couldn't get more than 17w with proper heat on fet and driver stage. I think that it could be 20-25 with harmonic content (power meter counts them as power too). Ofcourse I tried with 50ohm load, maybe someone tried with bad load or something?

[Maximus]

^^^This.
I managed to get almost 28w out of the supplied 15hvf1 after tweaking the bias and upping the psu to around 13v ish, but it got hot fairly quickly, plus was sprogging all over.
The optimum I found was 12w @ 12.5v (give a little due to the resistance of the wires and regultors etc) as it ran cool and harmonics were minimal.

When using it to drive the 150w amp at 3.5-4.5w depeing upon the freq, the signal was very clean.

[LAZY]

Here are the results of the tests Builtinashed suggested.

Power Pot Pins - Max 1.76 Min 0.3v. Both sides of R17 change when pot turned.
D1 Zenner Diode Output 0.3v

Q6 BC557 Transistor - Emitter=12.38v Base=1.77v Collector=1.76v
Q5 BC547 Transistor - Emitter=0.003v Base=0.37v Collector=11.8v
Q4 BC327 Transistor - Emitter=12.38v Base=11.8v Collector=0.04v

Leg 11 of TSA5511 (this is what turns on Q6?) Unlock 2.4v - Locked 1.77v

Do the results above help resolve the issue? Do the figures suggest that one / all of the above transistors are shagged?

Cheers. Lazy.

[Shedbuilt]

Here are the results of the tests Builtinashed suggested.

Power Pot Pins - Max 1.76 Min 0.3v. Both sides of R17 change when pot turned.
D1 Zenner Diode Output 0.3v

Q6 BC557 Transistor - Emitter=12.38v Base=1.77v Collector=1.76v
Q5 BC547 Transistor - Emitter=0.003v Base=0.37v Collector=11.8v
Q4 BC327 Transistor - Emitter=12.38v Base=11.8v Collector=0.04v

Leg 11 of TSA5511 (this is what turns on Q6?) Unlock 2.4v - Locked 1.77v

Do the results above help resolve the issue? Do the figures suggest that one / all of the above transistors are shagged?

Cheers. Lazy.

Q6 definitely looks to be blown. It's PNP, and shown in the diagram with emitter connected to the +ve rail, so the base should never be more than 0.7v below the +ve rail (so base at or near 12.38v when turned off, and should be between 11.7 and 11.8v when turned on. R25 pulls the base up to the +ve rail, for the switched off condition, and the PLL (yes, it's pin 11), must pull the base down, to switch it on. It looks from the above, that Q6 is open circuit on the base / emitter junction, and maybe short circuit base to collector.
Can't tell for sure whether Q4 and / or Q5 are ok. There's nothing which looks unreasonable, from the voltage readings, but with no voltage (or nowhere near what there should be), to the pot, Q5 is not getting enough base voltage to even start to turn on, and if Q5 is fully off, Q4 is fully off too. The voltages around them look about right for the turned off condition, and I think there's a fair chance that both are ok, but can't tell for sure what's going to happen when they start receiving the right conditions.

[LAZY]

Excellent. Thanks for all your input so far. I'll order some replacement parts and try again.

Cheers. Lazy.

[Shedbuilt]

No problem !
Since Q6 is just a switch, you could, if you are careful, and follow the right steps, temporarily bypass it, to test the rest of the circuit. Your choice of course. You would need to remove it before doing anything else; especially because it looks as if it may well have a base / collector short (which would otherwise send supply voltage back to the switching pin of the PLL chip). Once you removed Q6, you could (with the board powered down), take a connection directly from the +ve supply rail on the board, or to the positive wire from the board - as long as it's not carrying power at the time, to the "Maximum" end of the power pot (opposite end to the end which is grounded).
Make sure the pot is set to minimum, before you power the board up, and leave it there until the board is powered up, and locked. Minimum is effectively "Off" anyway, and I think you'll find a little bit of a dead band at the bottom end of the control. The wiper of the pot, will need to come up to something over 3V, before Q5 even starts to turn on. Based on some quick maths, and if my understanding of the regulator circuit is correct, I'd expect 4v on the wiper of the pot, to produce about 1.5v out (going out to the top end of R23 and R11, and off to the bias control circuit). With 8v on the wiper, I'd expect to see about 8v out, and with 10v on the wiper, I'd expect about 11.3v out.
Of course, once this is supplying voltage as it should, if everything else is in order, you should get some RF out. If you still can't get any output, it's likely to mean that one or both RF amplifiers are blown, but again, one thing at a time.
Having said that, the RF to the PLL comes from the output of the MSA device, so the fact that the PLL works, locks etc, is a reasonable indication that the first amplifier at least is alive.......

[MiXiN]

Have the replacement components arrived yet, Lazy?

With all the kind assistance on here, I don't think it'll be too long until you're up and running.

[LAZY]

Yeah, they're here ready to be installed. I was going to try Builtinashed's idea in the above post but decided to wait until I had a bit of time and do it properly.

No doubt I'll be posting up for help tomorrow at some point.

Lazy.

[MiXiN]

Good luck mate, and looking forward to seeing the finished product.

[LAZY]

IT'S ALIVE. ALIVE! It was Q6 the BC557 that was shagged. Now Just the tuning and canning to do. I've done some very quick power as its not mounted on a heatsink yet. For power testing i'm using a Zetagi HP500, which TBH is a pile of poo. At the edges of the band is where i'm seeing most power, not so much in the middle.

Can anyone recomend an accurate(ish) power meter for not silly money. I remember testing it with rigs some years ago and getting much lower figures than the rigs were rated for. (i know this is not unusual for pirate tx's) But my source at the time was good!

Next on the list is boxing up. It comes supplied with a metal tin. Do I also require a heatsink? (Any recomendations on where to purchase one) If so how do I go about attaching it? Do I need to cut a hole in the bottom of the tin and attach the Final directly to that or will attaching box directly to the heatsink suffice? I have some artic silver thermal paste, is it any good for this appication?

Directly translating from the guide I get this "Before the halves are soldered to the board , there must first be some holes drilled for the penetrations. Place three 1nF penetrations for the potentiometer and one for food. The audio input and RF output gain no transit ! The audio throughput can , eg . Made of a piece of rigid coax. The RF output can be directly soldered a piece of coax or SMA socket."

Ok so I drill holes for the potentiometer and the power and put 1nF capacitors in the holes? Is that correct?? What sort of capacitors? Here's a picture from the guide.




Thanks for all your help so far. It really is very much appreciated.

Lazy.

[Gum]

Nice work!

Might be a daft question, but did you calibrate your power meter using the knob on the front? Meters like that are usually OK at giving you relative measurements but you need to calibrate them more or less every time you use them, especially if you change frequency between measurements...

A word of warning about thermal paste, the silver and gold stuff is electrically conductive! Accidentally drop a blob on your board and it will short out everything it touches... and it can be a total cunt to get off the component side as the more you clean it, the further it works its way into all the gaps. If you do use thermal compound, do yourself a favour and use the white stuff which is non conductive.

You only need the tinyest amount mind.

[pjeva]

That's nice LAZY! Regarding that Zetagi, I think it covers up to 200MHz so i think it is not problem for it to handle FM band. Did you wind coils in LPF as it says in manual? If you have more power on the ends of the band, maybe it acts as band reject filter... Capacitors that you should put on the box are so called pass through capacitors (you can see them on pictures above or internet). If you didn't get them in kit, don't bother to buy them. Just pass some wire through the holes, it will work. For audio, you can use regular cheap rg58 and try to solder ground as close as possible to hot end. Regarding heatsink, ofcourse you need one. Without it, you can run it with 0.5w without damage... if you use metal box for whole device, you can use wall (bottom) of the box for heatsink. 6W will disipate heat that you can cool with piece of aluminium approx. 8" × 3" and 0.1" thick

[Shedbuilt]

Great news Lazy.
Although, like pjeva says, it will work without the feedthrough caps, I'd personally be inclined to use them. To some extent, it depends on the strength of the RF field. The feedthroughs will help to suppress RF picked up in the wiring. In the case of the pot, if it picks up a lot of RF in the wiring, it could make the regulation circuit go funny (affecting voltage to driver, and bias to final). It could also feed back into the RF circuitry, which might cause undesirable effects. Dutch RF shop sell them http://dutchrfshop.nl/componenten/conde ... r-1nf.html also on ebay http://www.ebay.co.uk/itm/25-x-Feedthro ... 27f80b03de
In this video - at about the 1 minute mark, you can get an idea of the size of heatsink used




I had a Zetagi meter - the same one I think. I used to find it reasonably ok, and as pjeva says, it is specified for 200MHz, and even if not totally accurate (which also depends on the SWR presented by the dummy load), it should be fairly consistent across the FM band. How much does it fall off in the middle of the band ?

Might be a daft question, but did you calibrate your power meter using the knob on the front?

The calibration knob is only in use when switched into "SWR" mode. You set the forward power to F.S.D, then check the reflected on the other meter. It's disabled when you use it as a power meter.

[LAZY]

Ok thanks guys. On the edges of the band I'm getting about 9w. In the middle I'm getting 3w. I can only test for a few seconds due to the lack of heatsink.

Next step cooling. I'll see if I can grab one cheap on ebay and start testing. Then I'll grab some feedthroughs and come back with an update and some more pics.

Cheers guys.

Lazy.