Acoustic 150 Question
- Thread starter CJ
- Start date
Help Support GroupDIY Audio Forum:
Because the audio signal is at the base of the power devices. The cap is shunting the AC 0V opposite end of the transformer windings.
The top one is bootstrapped to the audio output. So pull up drive is floating above the audio output. The pull down drive is relative to ground.
JR
PS: probably an old tube design that was transistorized.
The top one is bootstrapped to the audio output. So pull up drive is floating above the audio output. The pull down drive is relative to ground.
JR
PS: probably an old tube design that was transistorized.
CJ
Well-known member
thanks John!
Audio1Man
Well-known member
The caps keep the BIAS voltage stable and provide a 0 VAC return from the transformer drive as John said. The amplifier was not a redesign of a tube amplifier. The drive signal is push pull to the output transistors.
You may want to look how signed the schematic. Duke
You may want to look how signed the schematic. Duke
CJ
Well-known member
It may be a novel amp design but my speculation is that it looks like one foot standing in old vacuum tube topology, and the other foot in solid state.Audio1Man said:
I recall seeing an early solid state amp ('60s Knight Kit) that used a similar hybrid topology. It is an effective (but not cheap) way to deal with using only one polarity of output power transistor. In the bad old days we didn't have decent complementary power devices.
The cheaper approach was to use a quasi-complementary topology (small PNP in front of NPN power device forming a compound darlington), but that suffered from different speed zigging as zagging .
JR
Audio1Man
Well-known member
Hi John During my time @ Acoustic (8+ years) I designed many products. The first proto 150 amplifier was a great HiFi amp and it was "BAD SOUNDING" not enough" MUSICAL SOUND" many other colorful descriptions from marketing. A redesign to add back distortion products, increasing the output Z to let the speaker ring made the product viable. It was a new learning process for me. You would not like using this amplifier for music system.
CJ, Mike Botich was my artistic man and draftsman. If you want a service manual let me know.
Duke
CJ, Mike Botich was my artistic man and draftsman. If you want a service manual let me know.
Duke
Cool... always great to hear from the horses mouth.Audio1Man said:Hi John During my time @ Acoustic (8+ years) I designed many products. The first proto 150 amplifier was a great HiFi amp and it was "BAD SOUNDING" not enough" MUSICAL SOUND" many other colorful descriptions from marketing. A redesign to add back distortion products, increasing the output Z to let the speaker ring made the product viable. It was a new learning process for me. You would not like using this amplifier for music system.
CJ, Mike Botich was my artistic man and draftsman. If you want a service manual let me know.
Duke
At Peavey one of their many patents to emulate tube amp performance was a tricked up variable output impedance to generate "more" cabinet/driver interaction on demand. Generally undesirable for hifi amps, but useful in musical instrument amps.
IIRC Peavy even played with different output impedance for HF and/or LF.
JR
CJ
Well-known member
R307 to R306 sets the gain from Q310 input to driver transformer primary. About 468 for AC, unity for DC.
AC gain to speaker jack set by R310 to R306, about 100.
Gain through driver transformer (step-down) and final transistors is round about unity more or less roughly. So there's a bare 4:1 NFB around the driver transformer and power devices. Considering that is 2 or 3 heavy high-cuts together, 4:1 may be all they could take.
R309 C306 work at 100KHz, so can only be supersonic roll-off for loop stability.
There's no obvious way the current in Q302 is set. Actually I think it is set by drop in R311. Q301 is clearly flowing 1mA (680r and Q302 Vbe). This splits in R307 and R308. I can't make the numbers work out, but clearly if Q302 sucks on R311, the "40V" falls and Q301 is turned off, turning off Q302, suckage is reduced.
Why do you want to understand? Is it working or not?
AC gain to speaker jack set by R310 to R306, about 100.
Gain through driver transformer (step-down) and final transistors is round about unity more or less roughly. So there's a bare 4:1 NFB around the driver transformer and power devices. Considering that is 2 or 3 heavy high-cuts together, 4:1 may be all they could take.
R309 C306 work at 100KHz, so can only be supersonic roll-off for loop stability.
There's no obvious way the current in Q302 is set. Actually I think it is set by drop in R311. Q301 is clearly flowing 1mA (680r and Q302 Vbe). This splits in R307 and R308. I can't make the numbers work out, but clearly if Q302 sucks on R311, the "40V" falls and Q301 is turned off, turning off Q302, suckage is reduced.
Why do you want to understand? Is it working or not?
Audio1Man
Well-known member
Hi CJ
PRR has gone through the circuit. It has been 36+years since the development, The output stage has only a small amount of NFB, and this allows the output Z to be high. It allows the speakers (the alum dome) to ring about 2200Hz witch gives the "TONE/COLORING" that we were looking for. If this is a 13x product look @ the .1uf/22ohm and see if it is discolored. The early 13x amps were modified for INPUT BW. The input signal cable (cheep china types) had poor shielding and when the cord drooped in front of the speaker HF FEEDBACK burnt up the 22ohm. R106 / 10k was bypassed on the preamp with a cap (I think it was a few hundred pf).
The driver TX is bifilar wound for the base drive. I don't recall who made the TX, however look for the MFG code stamped on it.
You asked about the hum, check the 3900uf and the 250uf caps.
Good hunting Duke
PRR has gone through the circuit. It has been 36+years since the development, The output stage has only a small amount of NFB, and this allows the output Z to be high. It allows the speakers (the alum dome) to ring about 2200Hz witch gives the "TONE/COLORING" that we were looking for. If this is a 13x product look @ the .1uf/22ohm and see if it is discolored. The early 13x amps were modified for INPUT BW. The input signal cable (cheep china types) had poor shielding and when the cord drooped in front of the speaker HF FEEDBACK burnt up the 22ohm. R106 / 10k was bypassed on the preamp with a cap (I think it was a few hundred pf).
The driver TX is bifilar wound for the base drive. I don't recall who made the TX, however look for the MFG code stamped on it.
You asked about the hum, check the 3900uf and the 250uf caps.
Good hunting Duke
CJ
Well-known member
we had two problems, one problem was excess hum which has been fixed by replacing C307 250/50, we stuck a 220uf at 250 volts (because that is what we had) and the hum is gone, quiet as a mouse, both preamp inputs are working, it has a nice vibrato circuit, ok reverb,
the last concern is voltage droop on the other side of R311-560 ohms, it is getting hot and we have 25 volts instead of 40,
resistors have been checked ad are ok, maybe the transistor on the built in heat sink (Q302) is tired, the first transistor has been replaced
with an MPSA06 instead of the MPSA09 and seems to be ok, (difference is a higher current rating, 500ma as opposed to 50 ma for stock)
C301 has been changed to 2 uf 100V non polarized
main filter cap has been replaced along with the output cap, which is 5000uf now instead of 1000uf,
here is a voltage chart, maybe tweaking the 680 ohm Q301 collector resistor might help?
thanks for all the help!
the last concern is voltage droop on the other side of R311-560 ohms, it is getting hot and we have 25 volts instead of 40,
resistors have been checked ad are ok, maybe the transistor on the built in heat sink (Q302) is tired, the first transistor has been replaced
with an MPSA06 instead of the MPSA09 and seems to be ok, (difference is a higher current rating, 500ma as opposed to 50 ma for stock)
C301 has been changed to 2 uf 100V non polarized
main filter cap has been replaced along with the output cap, which is 5000uf now instead of 1000uf,
here is a voltage chart, maybe tweaking the 680 ohm Q301 collector resistor might help?
thanks for all the help!
Attachments
Gene Pink
Well-known member
There were millions of push-pull transistor outputs with transformer base drive sold, starting in the late '50's through the '60's, the ubiquitous portable 9V battery powered "Japanese 6-Transistor AM radio".JohnRoberts said:
A good representative schematic is here, all PNP germaniums:
http://www.angelfire.com/planet/funwithtransistors/AJ6-1.html
FWIW,
Gene
Don't forget car radios but they still used tubes through the 50s..Gene Pink said:
JR
I still can't make the numbers work out. (Overdue for my meds?)
But if the readings at Q302 are accurate, Vbe is 1.2V when it should be 0.7V.
Could a Darlington have slipped into the Q302 holes?
Could C303 be leaking? (After 35 years.... you think?)
Q301's hFE looks non-wrong. (About 330 which is plausible.)
But if the readings at Q302 are accurate, Vbe is 1.2V when it should be 0.7V.
Could a Darlington have slipped into the Q302 holes?
Could C303 be leaking? (After 35 years.... you think?)
Q301's hFE looks non-wrong. (About 330 which is plausible.)
Attachments
One of those two voltage nodes is a PS rail so may have ripple V, or other errors. More reliable when measured directly base-emitter, instead of separately.
When in doubt the schematic is probably wrong, thats how the ancients used to test us.
Your current analysis suggests current must be flowing into the pnp base (unlikely).
Relax... it's just a drill.
JR
When in doubt the schematic is probably wrong, thats how the ancients used to test us.
Your current analysis suggests current must be flowing into the pnp base (unlikely).
Relax... it's just a drill.
JR
> When in doubt the schematic is probably wrong
Yes; but R311 should not be sitting AT its rated dissipation.
After sleep, I agree that "1.2V" across Q302 B-E is probably a measurement artifact. I still would check what is in that socket, but a Darlington would not make the numbers like this. Assuming a simple Si transistor at Q302 we have 0.9mA in R305, 1.3mA in Q301, the 0.4mA difference would be plausible base current in Q302 which is nominally 62mA. That base current is a large fraction of drive current would be typical design from days when transistors were expensive, maximum current transfer and gain.
Yes; but R311 should not be sitting AT its rated dissipation.
After sleep, I agree that "1.2V" across Q302 B-E is probably a measurement artifact. I still would check what is in that socket, but a Darlington would not make the numbers like this. Assuming a simple Si transistor at Q302 we have 0.9mA in R305, 1.3mA in Q301, the 0.4mA difference would be plausible base current in Q302 which is nominally 62mA. That base current is a large fraction of drive current would be typical design from days when transistors were expensive, maximum current transfer and gain.
Ah. Wait. Q301 has 10K||22K= 6.8K in emitter and 680||VbeQ302 or say 500r in collector. 6800/500= 13.6 multiplication. Taking Vbe as 0.7V, 0.7V*13.6= 9.5V at Q301 emitter.
CJ measures 8.84V there. Working back, implies 0.65V for Q302 Vbe, which is also perfectly reasonable. Or various tolerances and drifts.
Take the split-diff, 9V at Q302 emitter. Then 9.6V at Q301 base. 470K:330K divider implies "40V" is 2.42 times higher, 23.2V at "40V". Q301 has maybe 4uA base current, which in 470K||330K= 194K is another 0.8V offset. 9.6V+0.8V= 10.4V, times 2.42 is 25V at "40V".
CJ confesses to replacing Q301. If he got a significantly higher hFE part than was supplied decades ago (even just two extremes of a typical 3:1 hFE range), then "40V" arrives at 29V. Still not jiving with factory numbers. If Q302 were also low-hFE in the prototype, we might come near 40V at "40V" even though the majority of production ran lower V here.
My vote is for R311 change to 680r or 750r at 10 Watts. Not obsess the "40V".
But still check Q302 for type and Vbe, and all e-caps for leakage.
CJ measures 8.84V there. Working back, implies 0.65V for Q302 Vbe, which is also perfectly reasonable. Or various tolerances and drifts.
Take the split-diff, 9V at Q302 emitter. Then 9.6V at Q301 base. 470K:330K divider implies "40V" is 2.42 times higher, 23.2V at "40V". Q301 has maybe 4uA base current, which in 470K||330K= 194K is another 0.8V offset. 9.6V+0.8V= 10.4V, times 2.42 is 25V at "40V".
CJ confesses to replacing Q301. If he got a significantly higher hFE part than was supplied decades ago (even just two extremes of a typical 3:1 hFE range), then "40V" arrives at 29V. Still not jiving with factory numbers. If Q302 were also low-hFE in the prototype, we might come near 40V at "40V" even though the majority of production ran lower V here.
My vote is for R311 change to 680r or 750r at 10 Watts. Not obsess the "40V".
But still check Q302 for type and Vbe, and all e-caps for leakage.
Audio1Man
Well-known member
The 40410 is PNP si T05 in a welded heat sink. -90V, 3w part. RCA family 40406-07-08-09-10 & 11. The voltage @ Q302 E should be about 30-40 volts. E-B = 0.6v Si junction. the Collector as I remember was 2-3 volt ? R307 provides some NFB. Replacing R311 with a higher value may cause loss of drive current. The replacement of Q301 with a new device I think was not needed. Q301 is running @ 1.3ma @ 30v is 40mw.
If the amplifier is working, why look @ making any changes. If changes are made the customer may reject your repair.
Duke
If the amplifier is working, why look @ making any changes. If changes are made the customer may reject your repair.
Duke
