I'm attempting to resuscitate an old Sherwood S8000 (schematic attached). After checking most of the tubes (everything save the 7868s and EM84), I find that several of the 6 12AX7s are showing bad. What's in there now is a hodgepodge: two marked "Made in West Germany" (Telefunken?), two marked "Sherwood Made in Holland", one Mullard, and one Sylvania.
According to a Hickock 600a, three are bad and one is weak:
Based on those readings, it looks like I may need to buy at least 4 new tubes.
The questions are: is there any optimal ordering or pairing of tubes? That is should I put the newest or strongest tubes in the first stage, last stage, etc.?
And I know this may be a religious subject, but what's a decent replacement tube? Most of the verbiage out there seems to be oriented towards guitar amps, but the consensus seems to be that the cheap Chinese tubes are dicey, the various New Sensor brands and JJs seem to be at one level, Genalex Gold Lion (gold pin) and Psvane a little better, and then you get to a level I'm not willing to go to for this application. I was kinda thinking the Gold Lion regular pins might be a decent choice, but I'm not sure where they would fall between JJ, EH, etc. and the gold pins.
Any suggestions would be appreciated.
According to a Hickock 600a, three are bad and one is weak:
Code:
Mfg. Tri 1 Gm Tri 2 Gm
phono 1st stage Made in West Germany good 940 good 910
phono 2nd stage Sylvania weak 780 replace 710
cathode follower Sherwood Made in Holland good 940 good 950
amplifier 1st stage Sherwood Made in Holland weak 850 good 900
amplifier (left channel) Mullard replace 570 replace 520
amplifier (right channel) Made in West Germany good 970 replace 715Based on those readings, it looks like I may need to buy at least 4 new tubes.
The questions are: is there any optimal ordering or pairing of tubes? That is should I put the newest or strongest tubes in the first stage, last stage, etc.?
And I know this may be a religious subject, but what's a decent replacement tube? Most of the verbiage out there seems to be oriented towards guitar amps, but the consensus seems to be that the cheap Chinese tubes are dicey, the various New Sensor brands and JJs seem to be at one level, Genalex Gold Lion (gold pin) and Psvane a little better, and then you get to a level I'm not willing to go to for this application. I was kinda thinking the Gold Lion regular pins might be a decent choice, but I'm not sure where they would fall between JJ, EH, etc. and the gold pins.
Any suggestions would be appreciated.
Last edited:
The strongest and best tubes should go in the phono stage (if you use it), especially the first tube.
Telefunkens would be good for that purpose, but you could easily sell those separately for significant money.
Both channels should have the same tube brand/types in the same positions throughout.
Telefunkens would be good for that purpose, but you could easily sell those separately for significant money.
Both channels should have the same tube brand/types in the same positions throughout.
Last edited:
Is this an S-8000 or an S-5000?
The S-8000 doesn't use EL84s but the early ones used 7591s and then switched over to the electrically equivalent 7868. They also didn't use 12AX7s. They did use the 6EU7 which is the 6 volt equivalent of the 12AX7.
The S-5000 did use the 12AX7s and they made versions with all three aforementioned output tubes including the 7189 which is the high powered version of the EL84.
Now, as far as the readings you are getting on your tubes, I have seen Mullards and Tele 12AX7s read 30% of their rated value and function just fine.
Run the Mullard in a circuit for a while and see if it tests higher. When they sit for a long time, the collected mass tends can settle on the cathodes and or the control grid and cause lower emission on test.
The S-8000 doesn't use EL84s but the early ones used 7591s and then switched over to the electrically equivalent 7868. They also didn't use 12AX7s. They did use the 6EU7 which is the 6 volt equivalent of the 12AX7.
The S-5000 did use the 12AX7s and they made versions with all three aforementioned output tubes including the 7189 which is the high powered version of the EL84.
Now, as far as the readings you are getting on your tubes, I have seen Mullards and Tele 12AX7s read 30% of their rated value and function just fine.
Run the Mullard in a circuit for a while and see if it tests higher. When they sit for a long time, the collected mass tends can settle on the cathodes and or the control grid and cause lower emission on test.
Thanks for the pointers.
According to the attached block diagram, it appears that there are two stages to the phono preamp, with one tube handling both stages for channel 1 and a second tube handing both stages for channel 2. However, the schematic (attached to the previous post) shows that one tube handles the first stage of both channels and the second tube handles the second stage of both channels.
If I understand this correctly (and giving the schematic preference), I would put the best tube (e.g. Telefunken) first and the second strongest tube next. The third tube is a cathode follower (one section for each channel), fourth is the first amplifier stage (one section for each channel), and the final two are the second amplifier stage (one tube for each channel).
So, if I want the two channels to be as identical as possible, I would match the two sections within each of the first four tubes, then match both sections of the last two tubes (tube 1, triode 1 matches tube 2, triode 1; tube 1 triode 2 matches tube 2, triode 2).
Not trying to be overly anal here, but this design doesn't appear to have any facility for adjusting the bias on the output tubes. They spec matched pairs for each channel, but given the fixed bias would it not be even better if the signal coming to the pair on channel 1 went through approximately the same path as the signal coming into channel 2? That is, would it be better if each stage were the same (or close) for each channel?
Thanks.
According to the attached block diagram, it appears that there are two stages to the phono preamp, with one tube handling both stages for channel 1 and a second tube handing both stages for channel 2. However, the schematic (attached to the previous post) shows that one tube handles the first stage of both channels and the second tube handles the second stage of both channels.
If I understand this correctly (and giving the schematic preference), I would put the best tube (e.g. Telefunken) first and the second strongest tube next. The third tube is a cathode follower (one section for each channel), fourth is the first amplifier stage (one section for each channel), and the final two are the second amplifier stage (one tube for each channel).
So, if I want the two channels to be as identical as possible, I would match the two sections within each of the first four tubes, then match both sections of the last two tubes (tube 1, triode 1 matches tube 2, triode 1; tube 1 triode 2 matches tube 2, triode 2).
Not trying to be overly anal here, but this design doesn't appear to have any facility for adjusting the bias on the output tubes. They spec matched pairs for each channel, but given the fixed bias would it not be even better if the signal coming to the pair on channel 1 went through approximately the same path as the signal coming into channel 2? That is, would it be better if each stage were the same (or close) for each channel?
Thanks.
Attachments
Sorry -- typo.
I meant EM84. OP has been corrected.I dunno. I got a chassis here with 6 12AX7s in it and a schematic to match.
@Konga Man
The triodes in 6EU7s and 12AX7s are the same. The 6EU7 has both a unique pin-out and a spiral wound hum bucking heater, which favor its selection in AC heated units. Except for a spiral wound, hum bucking heater, the 7025 is identical to a (folded heater) 12AX7.
The current production Sovtek 12AX7LPS is a genuine 7025 equivalent. LPS = long plate spiral. Properly culled 12AX7LPS specimens purchased from a reliable dealer, like Jim McShane, will serve quite well in the S8000. Pay the premium for phono grade tubes, if you use the phono section.
An examination of the 12AX7 version S8000 schematic shows DC heated preamp tubes. So, hum bucking is not essential. Buy good sounding, quiet, tubes. More than 1 variety of current production should be fine, without the $ outlay buying desirable OS tubes entails. FWIW, I agree with the avoid Chinese guano line of thought.
The triodes in 6EU7s and 12AX7s are the same. The 6EU7 has both a unique pin-out and a spiral wound hum bucking heater, which favor its selection in AC heated units. Except for a spiral wound, hum bucking heater, the 7025 is identical to a (folded heater) 12AX7.
The current production Sovtek 12AX7LPS is a genuine 7025 equivalent. LPS = long plate spiral. Properly culled 12AX7LPS specimens purchased from a reliable dealer, like Jim McShane, will serve quite well in the S8000. Pay the premium for phono grade tubes, if you use the phono section.
An examination of the 12AX7 version S8000 schematic shows DC heated preamp tubes. So, hum bucking is not essential. Buy good sounding, quiet, tubes. More than 1 variety of current production should be fine, without the $ outlay buying desirable OS tubes entails. FWIW, I agree with the avoid Chinese guano line of thought.
JJ ECC83/12AX7, long plate, short plate, and gold pin versions all have Spiral Filaments.
I get mine at Eurotubes.com. I just drive over and pick them up.
Eurotubes re-tests the tubes, matches them if you want, and one important test they do is for noise.
They do ship to some overseas countries.
I get mine at Eurotubes.com. I just drive over and pick them up.
Eurotubes re-tests the tubes, matches them if you want, and one important test they do is for noise.
They do ship to some overseas countries.
Last edited:
No. The amp hasn't been fired up in 30-35 years. It was retired then because it kept blowing the fuse. As such, I'm not giving it line voltage until I've done a sanity check, restuffed the cans, replaced the other electrolytics and (maybe) coupling caps, and tried it with 25VAC or so. Or I could plug it in and follow the smoke.
Or you could start figuring out why it blew the fuse. Standard procedure, thenNo. The amp hasn't been fired up in 30-35 years. It was retired then because it kept blowing the fuse. As such, I'm not giving it line voltage until I've done a sanity check, restuffed the cans, replaced the other electrolytics and (maybe) coupling caps, and tried it with 25VAC or so. Or I could plug it in and follow the smoke.
replacing whats needed.
To necro this thread...
It took a while to poke around inside to figure out what's going on -- partially because this isn't my day job, and partially because the schematic, parts list, and block diagram in the original manual do not always agree with each other or the wiring in the chassis.
The voltages from the PS seem to check out, save the preamp plate feed. Listed as 242V on the schematic, I get 215V at the cap. The observed voltages at the plates (pins 1 and 6 of V1-2) check out, though, so maybe it's another error in the schematic.
I've also measured every resistor in the unit. Some 47(!) are outside their listed tolerance; some are wildly so (e.g. 1M: 500K, 68K: 24K, 12K: 1.3, etc.). Is it normal to have that many resistors (about 1/3 of the total complement) drift out? That's a lot of tight-quarters soldering to replace them.
I also measured the plate current at each of the 4 output tubes (all Sylvania 7868; they may be original). Although the tubes test good on a Hickock 600A with very close Gm readings (5700-5900 mhos), measured currents vary greatly (6mA, 8.8mA, 8.5mA, 10.9mA). If a tube is moved from the inboard to the outboard position, the measured current follows the tube. Is there any way to interpret this other than the tubes are wildly mismatched?
Other than that, the readings at the preamp and amp tubes are generally on spec (V2 has low cathode readings (.64V vs. 1.72V); V3 has high plate voltage on triode 1 (220V vs. 180V)). However, several of the tuner tubes have wildly divergent readings. Given Sherwood's reputed penchant for making updates without changing the documentation, it's not clear if these are actual functional issues.
Any pointers or suggestions would be greatly appreciated.
It took a while to poke around inside to figure out what's going on -- partially because this isn't my day job, and partially because the schematic, parts list, and block diagram in the original manual do not always agree with each other or the wiring in the chassis.
The voltages from the PS seem to check out, save the preamp plate feed. Listed as 242V on the schematic, I get 215V at the cap. The observed voltages at the plates (pins 1 and 6 of V1-2) check out, though, so maybe it's another error in the schematic.
I've also measured every resistor in the unit. Some 47(!) are outside their listed tolerance; some are wildly so (e.g. 1M: 500K, 68K: 24K, 12K: 1.3, etc.). Is it normal to have that many resistors (about 1/3 of the total complement) drift out? That's a lot of tight-quarters soldering to replace them.
I also measured the plate current at each of the 4 output tubes (all Sylvania 7868; they may be original). Although the tubes test good on a Hickock 600A with very close Gm readings (5700-5900 mhos), measured currents vary greatly (6mA, 8.8mA, 8.5mA, 10.9mA). If a tube is moved from the inboard to the outboard position, the measured current follows the tube. Is there any way to interpret this other than the tubes are wildly mismatched?
Other than that, the readings at the preamp and amp tubes are generally on spec (V2 has low cathode readings (.64V vs. 1.72V); V3 has high plate voltage on triode 1 (220V vs. 180V)). However, several of the tuner tubes have wildly divergent readings. Given Sherwood's reputed penchant for making updates without changing the documentation, it's not clear if these are actual functional issues.
Any pointers or suggestions would be greatly appreciated.
Replace any doubtful resistors with 1% metal film types. It's a good idea to do them all, if you'll keep it.
Tubes can differ in plate DC voltages by +/- 10% or more, and be fine. Clean the tube pins and sockets.
Also check the power supply dropping resistors, these can drift a lot, and metal film is not wasted on those.
Check for coupling capacitor leakage by measuring the following DC grid voltage, should be zero in most cases.
Tubes can differ in plate DC voltages by +/- 10% or more, and be fine. Clean the tube pins and sockets.
Also check the power supply dropping resistors, these can drift a lot, and metal film is not wasted on those.
Check for coupling capacitor leakage by measuring the following DC grid voltage, should be zero in most cases.
Last edited:
ID take wire wounds for the power supply RC filter resistors. because metal film is not really designed for peaks. But that is just in applications where there is a filter capacitor to ground.
Metal films are excellent for everything else. If there are any carbon comp grid stoppers leave them in place
Metal films are excellent for everything else. If there are any carbon comp grid stoppers leave them in place
Not "wildly mismatched". But all quite low! I'd expect 20mA-40mA depending. Altho 7189 does have a low idle current suggestion.
If you bought this amp as a studio reference tool, it wants a better output set. But as a recreational amplifier, I would not be over-concerned. If this was 1970 I might dial-up the bias and idle current to get away from crossover. But my experience with a similar Fisher says "best" sound is at 150%-200% of rated dissipation, and they cost a little more today than back then.
All the pins and sockets have been cleaned.
After bumping the input voltage from 110V to the specified 117V, the output 7868s now show plate currents of 33.3 mA, 50.1 mA, 26.5 mA, and 45.2 mA. There's no provision in the design for adjusting the bias (cathodes directly to ground, grids from a common feed with fixed resistors). The trouble with sourcing two matched pairs or a matched quad is that the AFAIK the only new 7868s are the EH tubes with the broomstick pins, and I don't want to stretch out the sockets. Equally discouraging, the cost of appropriate NOS or good used vintage tubes is prohibitive.
As befits its age, the overwhelming majority of resistors are AB-style carbon comp. There are a few WW for the high-draw RC filters. There are also 6 resistors of unknown (to me) type: light brown with rounded ends, used as plate resistors in the two phono preamp 12AX7s and as both plate and cathode load resistors on the amplifier sections of the two 12AX7s in the amplifier circuit. They look similar to modern film resistors (but a little larger and rougher), but I'm not sure what they might have been ~60 years ago. I had intended to replace like with like, on the understanding that new CC resistors are better (i.e. more stable) than their predecessors and (for OCD purposes) to keep the original look.
I measured DC grid voltages on all 6 12AX7s. Readings were negligible (e.g. .001V) on both sections of the two tubes in the phono preamp and the first amplifier tube. For the cathode follower (between the phono preamp and amplifier) grid voltage was 27V on pin 2 and 34V on pin 7. For the two tubes in the amplifier circuit, there was no DC voltage on the grids in the amplifier sections, but 19.5V and 17.7V at the grids of the phase inverter sections. Although those readings seem high, they all correspond to sections which are taking output from the cathode rather than plate. I figured that might be the explanation. Is that correct?
After bumping the input voltage from 110V to the specified 117V, the output 7868s now show plate currents of 33.3 mA, 50.1 mA, 26.5 mA, and 45.2 mA. There's no provision in the design for adjusting the bias (cathodes directly to ground, grids from a common feed with fixed resistors). The trouble with sourcing two matched pairs or a matched quad is that the AFAIK the only new 7868s are the EH tubes with the broomstick pins, and I don't want to stretch out the sockets. Equally discouraging, the cost of appropriate NOS or good used vintage tubes is prohibitive.
As befits its age, the overwhelming majority of resistors are AB-style carbon comp. There are a few WW for the high-draw RC filters. There are also 6 resistors of unknown (to me) type: light brown with rounded ends, used as plate resistors in the two phono preamp 12AX7s and as both plate and cathode load resistors on the amplifier sections of the two 12AX7s in the amplifier circuit. They look similar to modern film resistors (but a little larger and rougher), but I'm not sure what they might have been ~60 years ago. I had intended to replace like with like, on the understanding that new CC resistors are better (i.e. more stable) than their predecessors and (for OCD purposes) to keep the original look.
I measured DC grid voltages on all 6 12AX7s. Readings were negligible (e.g. .001V) on both sections of the two tubes in the phono preamp and the first amplifier tube. For the cathode follower (between the phono preamp and amplifier) grid voltage was 27V on pin 2 and 34V on pin 7. For the two tubes in the amplifier circuit, there was no DC voltage on the grids in the amplifier sections, but 19.5V and 17.7V at the grids of the phase inverter sections. Although those readings seem high, they all correspond to sections which are taking output from the cathode rather than plate. I figured that might be the explanation. Is that correct?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.