79 Ampeg SVT Volume Drop

[numberfourthree]

If it was the filter caps with the symptoms that you describe, the filter cap(s) would be HOT to the touch after a few minutes of operation (prior to bursting).

The resistance values on the OT look high to me.

Just ran them again to check and no heat from the caps that was noticeable. Run about 60vac into the amp though and with the limiter.

Any other recommended tests for the OT? That has been my suspicion as it's already a replacement and oddly I've seen several other vintage SVTs with replaced OTs for sale.

 

[Wasnex]

1st issue is V9 that flashed a few weeks ago is getting very hot, while v8-V4 stay cool. I have a spare Svetlana 6550 which solves the issue but then that tube stays cool while the rest of the 6550s get warm. Likely a short in the GE 6550, sadly. Will test tubes later this week.

All this says is the tubes are not matched. As tubes age their ability to emit electrons goes down. However it is possible they may drift hot or cold. The reason is because the control grid's ability to hold the electrons back can also degrade. Since V9 is hot with a GE and cold with the Svetlana, the grid network may be fine. If the other 5 tubes are pulling the same current, put one of them in V9 and move the Svetlana to fill the empty socket. If V9 then pulls about the same current as the other 5 matched tubes, you probably just need a 6th matching tube. I would probably buy a matched trio and keep the two good tubes as spares.

I have noticed a grounding issue. The CE Can Caps ground lugs are connected to its case so using those lugs as ground bus creates continuity btwn the can exterior - possible ground loop to entire chassis and exterior of trannies. Perhaps this is the short the Limiter is seeing?

The capacitor cases are supposed to be grounded. As I showed in post 90, the chassis ground for the power amp originates on the back panel at J1. This ground is distributed to the PC power board on a wire, then it travels from one side of the PC power board to the other on a solder trace. Various components are connected to the solder trace. On the other side of the PC power board, a wire distributes the ground to the capacitor network. So yes the capacitor cases are grounded through this path. We had you remove the metal mounts so the capacitor cases are not also grounded directly to the chassis.

A wire distributes the ground from the capacitor network back to the Power PC card near the HT diode bridge rectifier. Various resistors are also connected to ground in this part of the Power PC card, but it does not loop back to the chassis ground, or any other part of the amp for that matter.

Often the ground from the AC line is connected to one of the transformer lugs. But it may be attached to the chassis in another manner.

The vintage SVT does have ground loops in the preamp. But unless something has been wired wrong in the Power Amp, it should not have any ground loops there.

The way the lightbulb limiter is working does not indicate a sort. When you put the amp in standby the bulb dims. When you adjust the bias, the bulb changes intensity.

Both of my vintage SVTs pull almost 300W at idle. So a 150W bulb should light up pretty well, but still drop some voltage.

When you finally fix the amp, make sure and cool down the bias before you plug the amp back into 120V.

Fixing the ground loops in the preamp requires cutting solder traces and adding some new connections. Beyond my ability. The main chassis ground for the preamp also originates at J1 on the back panel.

 

[numberfourthree]

Thanks to both of you for your continued advice. Luckily pre-amp doesn’t seem to be much of an issue.
So it seems primary resistance on the OT should be closer to 1.8k. It seems the Heyboer replacements would be a good fit over the expense of a Mercury. Any experience with replacement OT?

 

[agedhorse]

I don’t think you are talking about dc resistance.

 

[numberfourthree]

I don’t think you are talking about dc resistance.

Ah yes Impedance measurement on the OT should be around 1850ohms. So little info on these transformers out there.

 

[Wasnex]

Not sure if this is textbook perfect, but it may be helpful:



At the end he calculates the impedance. I believe he is using this relationship:

Specifically, Zp= [(Vp/Vs)^2] x Zs is the formula he is using when he calculates (30^2) x 8=7.2k.

Do you know that expected plate load of an SVT OT? If you no, the calculation is not useful.

Also, not sure if this is the correct plate load calculation for class AB push pull, since only half the tubes are assumed working at a given time.

 

[agedhorse]

At this point you can ignore the impedance measurement since it's not even coming into play with no signal.

What is the DCR of the primary SUPPPOSED to be on this transformer? I would expect somewhere between 10 and 25 ohms.

 

[numberfourthree]

There is so little info on SVT transformers, oddly. Even those who make replicas, don't provide much info at all. So I don't have expected plate load or DCR.
Schematic says 34.6Vac should be load on secondaries from Orange to Blue. I've only been able to measure very low readings. Seems plenty of power is getting to the primary side but little on the 2nd.

 

[agedhorse]

I wonder if the wrong transformer was used? If the DCR is really as high as you measured, the transform ratio may be very far off too.

You need to get more information on the OT first IMO

 

[Wasnex]

I wonder if the wrong transformer was used? If the DCR is really as high as you measured, the transform ratio may be very far off too.

You need to get more information on the OT first IMO

My understanding is the amp worked fine and made good power with this OT. Then somethings changed and it became weak and started intermittently blowing fuses.

IMHO the first thing I would suspect is the tubes. They may measure good on a tester, but tester dont typically run at real world SVT voltages.

I don't have any idea what the resistance readings for the OT should be.

A source on music-electronics-forum.com said the primary impedance spec is 1850 at 2 and 4 ohms. Ampeg SVT Specs - Music Electronics Forum

Cerrem posted that the primary is 1.6k at 4 ohms and 1.75k at 2 ohm. Ampeg SVT-CL question

If the formula I post earlier is the correct formula, I believe that makes the voltage ratio 21.5:1 at 4 ohms

[sq rt (1850/4)] = 21.5058
​

Both MM and Heyboer make specific OTs for vintage SVTs.

 

[beans-on-toast]

Reading the DCR of a transformer will tell you if there is an open or short but it also needs to be tested to determine if cracks in the wire coatings (electrical insulation) are leading to shorts when power is applied, a hipot test does that. Most can’t do that at home as it requires a special high voltage unit.

Sending the transformer for testing to Mercury or Heyboer would be expensive as you’d need to cover shipping both ways in addition to their charges. Heyboer will rewind your vintage transformer, unwinding it and sending them the plates and other parts reduces your cost. Expect a long wait.

Mercury and Hammond has sent me the DCR’s for their products. If you explain that you think that you have a faulty transformer and are considering buying a replacement from them, they might send you those specs.

 

[numberfourthree]

Yes I have emails out to heyboer and mercury for more detailed specs on the OTs.
As you said Wasnex, the amp worked fine for years. Blew a fuse and when replaced was weak with distortion. Changed out V1 and amp came back for a minute and then next time I played it went weak again. OT is a replacement so I would just buy a new one if needed.

 

[agedhorse]

A DCR that high is really suspect, but without the data it’s just suspect and nothing more.

 

[beans-on-toast]

Bill Hughes on the two parallel secondaries, he said that he had to get creative to implement the design. It made it easier to design and build the way that it was done. There are interleaved windings. There was the possibility of doing something with the secondaries in the future.

 

[numberfourthree]

Phil from Heyboer got back to me with some details:

3 primary interleave splits 400-800-400 turns of #25 ga magnet wire, are between 4 secondary splits of 54-tap-27 turns of 23ga Trifilar ( 3x23ga). These 4 sections are the same, all in parallel then split out as two and two windings. Secondary 1 and 4 are a pair and Secondary 2 and 3 are another pair.

Primary = 1800 Secondary = 2 - 4.5 ohm

He does not have DCR figures. 1600 turns on the primary total and 54-27 (81 turns) on secondary X 4 two per pair in parallel.

Turns ratio is 29.62:1 on the 2 ohm and 19.75:1 on the 4.5 ohm

They have a modified version on file as well with bifilar wind vs trifilar of slightly larger wire. It was a mod to make things a bit cheaper as it’s not as difficult to wind 2 side by side as it is 3 and keep it nice and flat.
​

It is not so easy to wind these as the originals were done. Keep in mind, this is a repro transformer and not necessarily akin to originals but still need DCR to resolve the high ohm rating btwn primaries on mine.

No word from Mercury Magnetics.

 

[numberfourthree]

Almost a year since this project began. Of course, work life and other projects got in the way. I’ve occasionally put more work in the SVT, checked voltages, etc. A friend suggested checking out the external speaker jack, standby and power switches and the umbilical. Did not see any faults on the switches or jack but a good clean and reflow of solder while in there seemed worthwhile. Checked preamp, reslfowed solder, replaced any out of spec caps, resistors.

Blammo, amp was back to life. I can’t tell you have fantastic is was to hear the amp rattle the floor and ceiling again.
A bit more testing, Channel 1 is great - back from the dead, Channel 2 still displays the low power/volume issue.
My suspicions now center on preamp due to channel inconsistency. More so to the Molex cable which over time has been bent and frayed over time.

Has anyone rebuilt the molex? I saw @beans-on-toast mentioned attempted to replace the cabling per a ground issue. I’ve cleaned and retensioned connections but the whole umbilical cord seems a like design out of convenience rather than performance, reliability.

 

[Wasnex]

Almost a year since this project began. Of course, work life and other projects got in the way. I’ve occasionally put more work in the SVT, checked voltages, etc. A friend suggested checking out the external speaker jack, standby and power switches and the umbilical. Did not see any faults on the switches or jack but a good clean and reflow of solder while in there seemed worthwhile. Checked preamp, reslfowed solder, replaced any out of spec caps, resistors.

Blammo, amp was back to life. I can’t tell you have fantastic is was to hear the amp rattle the floor and ceiling again.
A bit more testing, Channel 1 is great - back from the dead, Channel 2 still displays the low power/volume issue.
My suspicions now center on preamp due to channel inconsistency. More so to the Molex cable which over time has been bent and frayed over time.

Has anyone rebuilt the molex? I saw @beans-on-toast mentioned attempted to replace the cabling per a ground issue. I’ve cleaned and retensioned connections but the whole umbilical cord seems a like design out of convenience rather than performance, reliability.

Good news!

Channel 2 tends to have less gain than channel 1. The amount of variation depends on the era. I have a 69 and the difference is pretty big. I also have an 87, and the difference is relatively small. I don't know what would be considered normal for a 79.

One difference for blackline SVT is R49 in the power supply was a lower value (4.7K 15W VS 8.2K 15W. Reducing the value increases voltage at node B, which is the preamp supply.

The parts for the molex connectors should still be available; the sockets and pins can be purchase separately. I haven't completely redone an umbilical, but I resoldered a couple of connection and I also ran a separate shielded cable for the signal wire in 2 of the 3 SVTs I have owned. It did not seem to help. Keep in mind the pin connections vary somewhat by era.

Beans posted the part numbers the molex connectors years ago, but they have all been superseded. You may be able to google and cross reference.

I believe these are Bean's hand-written notes.

 

[agedhorse]

Almost a year since this project began. Of course, work life and other projects got in the way. I’ve occasionally put more work in the SVT, checked voltages, etc. A friend suggested checking out the external speaker jack, standby and power switches and the umbilical. Did not see any faults on the switches or jack but a good clean and reflow of solder while in there seemed worthwhile. Checked preamp, reslfowed solder, replaced any out of spec caps, resistors.

Blammo, amp was back to life. I can’t tell you have fantastic is was to hear the amp rattle the floor and ceiling again.
A bit more testing, Channel 1 is great - back from the dead, Channel 2 still displays the low power/volume issue.
My suspicions now center on preamp due to channel inconsistency. More so to the Molex cable which over time has been bent and frayed over time.

Has anyone rebuilt the molex? I saw @beans-on-toast mentioned attempted to replace the cabling per a ground issue. I’ve cleaned and retensioned connections but the whole umbilical cord seems a like design out of convenience rather than performance, reliability.

I have used MTA style connectors for decades without any reliability issues, done right is more reliable and higher performance than other methods.

If you can't identify the bad connection on the header, than it's likely not that. Guessing isn't going to get you very far if you are wrong.

Rebuilding or reterminating a Molex header just requires the proper tooling and parts.

 

[Wasnex]

My suspicions now center on preamp due to channel inconsistency. More so to the Molex cable which over time has been bent and frayed over time.

Channel 1 and Channel 2 are mixed on the input to V4. V4 provides the signal to the umbilical. So if channel 1 works fine the signal path through the umbilical should be fine.

Several wires in the umbilical loop in and out of the preamp chassis without even going to the preamp circuit, AC power switch, Polarity Switch.

I don't believe the fan wiring goes through the umbilical, but it is terminated on the Molex.

 

[beans-on-toast]

Almost a year since this project began. Of course, work life and other projects got in the way. I’ve occasionally put more work in the SVT, checked voltages, etc. A friend suggested checking out the external speaker jack, standby and power switches and the umbilical. Did not see any faults on the switches or jack but a good clean and reflow of solder while in there seemed worthwhile. Checked preamp, reslfowed solder, replaced any out of spec caps, resistors.

Blammo, amp was back to life. I can’t tell you have fantastic is was to hear the amp rattle the floor and ceiling again.
A bit more testing, Channel 1 is great - back from the dead, Channel 2 still displays the low power/volume issue.
My suspicions now center on preamp due to channel inconsistency. More so to the Molex cable which over time has been bent and frayed over time.

Has anyone rebuilt the molex? I saw @beans-on-toast mentioned attempted to replace the cabling per a ground issue. I’ve cleaned and retensioned connections but the whole umbilical cord seems a like design out of convenience rather than performance, reliability.

Channel 1 and channel 2 are mixed in the preamp. There isn’t a channel 1 and a channel 2 signal cables going to the power amp. If channel 1 is working the molex is doing it’s job. I like to replace the shielded cable with a new one. The old ones tend to develop openings in the shield at kinks which lets noise in. The shield can also rust and break, again an entry point for noise.

Yes, the umbilical cable could be done better. The bundle carries signal and power in close proximity. Some people take the shielded cable carrying the preamp signal out of the bundle, running beside it so there’s some separation. In later revisions such as the VR and CL there are two bundled cables, one one each side of the preamp chassis. They keep the small signals away from the high voltage ones.