the Poor Man 660 support thread

[lolo-m]

1°) A transformer hasn't got any impedance.

- But a transformer is optimized to work at a certain impedance - thats why we have a 600/600 and a 10K/10K.
- My understanding of this is that the DCR will be very different between the 2 transformers in the above example.

DCR, inductance, capacitance... A lot of parameters are different between each type, and you're right transformers are optimized for a certain load  .
But good transformers can be used reversed with excellent results aswell with unproper loads... There are some use of transformer like this in the real Fairchild (input trannie of the SCamp: reversed with unmatched loads between primary and secondary...).

 

[emrr]

PM660 input impedance is 1800 ohms.  It's defined by the 30K attenuator.   If you have a 10K:600 in front of that (between your other comp and the PM660), then your other comp is seeing roughly 30K as it's load.    You should not see any loss because of that, unless you are using a single 12AX7 (or similar) as a line output stage.   I think you have another problem. 

(30K actual) 10K:600 / 600:10K / 30K attenuator

If you are seeing 7 dB loss, you could say that your other comp has an output impedance of 30K (6dB power matching loss), and the other 1 dB is accounted for by the quoted transmission loss of the Edcors. 

Consider it again backwards:

30K attenuator driven by 10K:600 driven by 600:10K = 30K load presented to source.

The 10K:600 / 600:10K cancel one another, minus transmission losses. 
The 10K:600 in between units only helps to lower the output Z of the source device; also meaning it eases the load on the source device.

or:

30K attenuator connected to 10K:600.  30K is 3x10K spec.  So the reflected impedance is likewise 3x600=1800. 

A matched load causes a 6 dB loss.  Drive the PM with a 1800 ohm source Z, and there will be a 6 dB loss.  In the case described, using a 10k:600 between device and PM, it would require a 30K source Z to see a 6 dB loss.    To ignore math, and go with gut and observation, even a 600 ohm source driving an 1800 ohm load should only see a tiny loss; a dB or so. 

Any self respecting piece of reasonably modern pro equipment will have an output impedance of under 200 ohms; even the majority of tube preamps with 600 ohm ratings.  This is due to application of negative feedback in the vast majority of circuit types.  Many times the defining source Z comes from build-out resistances within the output stage, applied to prevent short circuit conditions.  Many devices will have output impedances well under 100 ohms.  So, very few things will struggle significantly driving an 1800 ohm load. 

As stated a billion times in a billion threads, rated transformer Z's suggest the inductance spec, which varies depending on the target Z's the transformer is designed to mate.  So long as you drive with a source equal to, or lower than, you will meet frequency spec.  That's the simple version. 

 

[Holger]

I've spend some time today to fix this problem:

I've moved the regs on the back of the case, seems to be more reliable now.
I removed one  0R22 resistors on the slow start heater PCB, now I have 6.3 V for the heaters (had 5.8V before).

I furthermore replaced the 6BZ8 with russian 6N5P, works good.
Regarding gain and compression the channels are close together now, if not matched (some 1/10 dB difference).
I did not notice any bass roll off, but on one channel I have a roll off starting at ~5-6 kHz with the -3 dB point at 18kHz.
The other channel is a little bit better. This could be improved, even if I don't know how. Maybe I'll try to exchange some tubes again.
I've played a bit with the compressor, it seems to be a good one for acoustic sources like piano, bass and guitar, maybe vocals. The differences between the six time constant positions are a bit subtle, but I can clearly hear differences.
I don't see it as my preferred mastering/summing buss compressor, but I like it very much on some material.

 

[lewilson]

Try a .047 film cap between the first tube stage cathodes it should bring the top end up a bit.

 

[Kingston]

I've moved the regs on the back of the case, seems to be more reliable now.

I'm surprised that works at all. From my experience with the LM388 in the conditions of poorman, a mere case backside is woefully inadequate.

I've used this type regulator in similar conditions with poorman and a dual LA2A (dropping 3V with 3,3A... deadly hot conditions). Using cooling blocks from dead ATX PSU's is *barely* enough. In fact, with the 3,3A case I will have to install at least a double sized massive block for security reasons. The regulator heat protection circuit will latch on after a few hours, and it would kill all the tubes if left unnoticed.

another tip, those 1000uF heater caps are also woefully inadequate in poorman conditions. You won't have anywhere near DC voltages with them. It needs 4X 4700-10000uF. Too bad the PCB is undersized for this, and 4X 10000uF caps dangling sideways isn't exactly pretty.

 

[Holger]

@lewilson: I'll try that one day, thanks.
@Kingston: I will observe the temparature issue over the next weeks. This beast gets really hot.
Maybe I'll go for extra aluminium heatsinks.

 

[Moby]

another tip, those 1000uF heater caps are also woefully inadequate in poorman conditions. You won't have anywhere near DC voltages with them. It needs 4X 4700-10000uF. Too bad the PCB is undersized for this, and 4X 10000uF caps dangling sideways isn't exactly pretty.

Exactly. That's why I designed custom PCB's for my PM, with plenty of space for 4x10000uF 

 

[TheGuitarist]

Hey just wondering what people are using for a couple of parts.

In particular, those who are using a rotary switch for the bypass.

Those who are using a rotary switch for the 6 TC positions.

And also the power indication LED? Lamp? The big one that is around 18mm to fit Purushas case. I've searched farnell mouser and jaycar and just can't find anything.

So far for the bypass i was just going to find any 2 position, or if i had to go more, just block the switch from going more than 2 positions. And for the attack i was going to use a 6pos 2 deck (i'll be using a scamp booster) or 12 pos 2 deck 30degree switch because i can't find anything else that will line up with the printed numbers... which doesn't concern me too much but it'd be nice. Seems like a few people have done it on here with purushas cases.

 

[skal1]

PSU



What do Q5, Q4,Q3 Q2 and Q1 do in the power suppy  i am having a problem wit this thing.

When the 660 boards are connected to the psu voltage drops to about 50v , i have changed all the transistors and irfs about 5 times now same results are the same , is their any reason way i can not get ride of those transistors and just drop the  voltages over some resistors?

skal1

 

[Moby]

PSU

Did U checked the zeners?

 

[skal1]

PSU

Did U checked the zeners?

Yes all checked and replaced


skal1

 

[Moby]

I have no idea.. to be honest I don't have any problems with PSU, but did you checked consumption of the PM boards, can be that some of the tubes are bad. For example on 136V rail U have to expect something around 10ma for mono or 20ma's for stereo PM. Check that first. Or make a shunt resistor to simulate that on PSU board and if there is no drop your PSU is OK.

 

[elektrovolt]

What do Q5, Q4,Q3 Q2 and Q1 do in the power suppy  i am having a problem wit this thing.
When the 660 boards are connected to the psu voltage drops to about 50v , i have changed all the transistors and irfs about 5 times now same results are the same
skal1

funny... I am having the same problem. replaced the IRFs, zeners and resistors several times. once after replacing the IRFs the voltages were OK. but after like 20 minutes the voltages dropped again (meter dropped, audio crapped out until no audio...

the problem seems present with both my comp boards, and also with my newly stuffed 3rd board. something stresses the PSU, but I don't have a clue what !

 

[skal1]

i had this problem before as well i do not think is the transformer but the circuit around q1 q2 q3 and  q4 .

Can a kind sole explain how and what this part of the circuit is actually doing.


skal1

 

[Moby]

I really think that U guys don't have a problem with PSU. Of course if PSU is workin in stand alone.  First , double check your wire connections, than audio transformers, especially T2 , T4 because it's in  the line with PSU. At last, try swaping your tubes to be sure that's not an issue.

 

[elektrovolt]

Moby,

wiring is OK, tried several tubes. how do i check the transformers then ? measuring resistance and check for leaking from primary to secundary ?

 

[skal1]

@Moby

my pm was working with the tubes what i had , i accidentally drop the psu in my new  case after testing it and that blow the irfs and mpsa.



Can a kind sole please explain how and what this part of the circuit is actually doing. q1 ,2,3 and 4.


skal1

 

[skal1]

Hi lads



can someone supply me with the ref voltages for the psu board for c1 c3  q1 q2 q3 q4 and d8 , with the pm660 boads connect and  without thanks .

skal1

 

[Moby]

Hi lads



can someone supply me with the ref voltages for the psu board for q1 q2 q3 q4 and d8 , thanks .

skal1

No prob, much easier than explaining point to point ) Here we go, this is the situation with my psu , I have a 300v after the bridge and 220u (c1)

Q1-E298.8V,B298.2V,C254V
Q2-C254V,B198V,E197.6
Q3-C197.6V,B100.5V, E99.9V

Q4-G254V,D299.4,S251.2
Q5-G137.4V.D251V,S135V

 

[skal1]

@moby
ok i can see a major difference in the voltage after the bridge  mine is reading about 451v thats over 150v more , whats wrong here ? have my caps gone bad.

thanks for the refs are they wit the 660 boards connect or without.


skal1