OT neg feedback winding

Paul Lebow

Member
2019-12-05 5:44 pm
Replaced a blown OT on an old Ampeg M15 30w, two 6L6's. Even though the schematic shows that the negative feedback from the OT is the same lead that goes to the speaker, it turns out the actual OT uses an additional winding for feedback.

Checking around in another forum the claim is that the original OT "OT-213" is a 6.5K 8/16. My replacement is a Hammond Vibrolux type 4K 4/8/16. Seems to work fine. My question is:

Why use an additional winding for the negative feedback? ( it is nice having the extra lead but...)?

Should I assume that the original amp used the 16 ohm tap for feedback and I should also use that winding in my repaired amp?
 

Paul Lebow

Member
2019-12-05 5:44 pm
Ampeg M15 negative feedback quandary

@leadbelly. Well it does seem that I did jump to a conclusion. For the stock OT-213 OT there is actually a separate winding with 2 leads. The schematic just shows a tie from the speaker output directly back to cathode of the inverter stage[IMGDEAD]https://el34world.com/charts/Schematics/files/Ampeg/Ampeg_m15.pdf[/IMGDEAD]
There is no connection between these leads and the other output leads but about 1.2 ohm between them. I believe one lead went to ground and the other through the feedback resistor. Seems that low resistance , few windings (if not shorted), would provide a very weak feedback signal.

The "bad" output lead has 1.2K between the c.t. while the "good" output lead has 120 ohms to the c.t. The amp had very weak output before I switched to a 4K Hammond 6L6 OT.

There was another anomaly with this amp compared to the schematic, the "stock" plate resistors of the inverter stage were 470k not the 120K on the schematic and the voltage ran way higher than that spec'd on the schematic (attached) I replaced with 120k's. The stock resistors seemed original, still had the adhesive on top of them.

I spoke with transformer vendor who believes the OT-213 is a predecessor of the OT-215 which from scanty information only has an 8 and 16 ohm tap. Right now I am feeding back from the speaker output as per schematic.
 

Attachments

  • Ampeg M15 schematic.pdf
    1.8 MB · Views: 19

Paul Lebow

Member
2019-12-05 5:44 pm
On stock (as I bought it 20 years ago)There are 4 output wires. One pair to speaker (yellow and black), other pair to feedback and ground (green and black)

Speaker pair - black to yellow wire - 0.7ohm
Feedback pair green to black wires - 1.0 ohm
Speaker black to feedback black wire - 0 ohms - so they are common
Speaker yellow to feedback green - 0.1 ohms (mysterious)

There are 3 input wires blue, red, brown.
blue to V4 6L6 plate
brown to V5 6L6 plate
red to B+

blue to red - 0.123K
brown to red - 1.18k (bad winding)
brown to blue - 1.6k

On my Hammond 4k OT I connect the common red input to B+ and the two other leads to V4 and V5 plates. Connect in-phase 8 ohm output to speaker and feedback resistor to inverter cathode. Connect other wire to ground and speaker ground. Tried connecting 16 ohm output to feedback instead - not much difference in tone.
 
There may be different versions, not well documented.
Mr Joe Piazza´s excellent work redrawing old fuzzy schematics may very well not match the version you have.

That said, circuits are not etched on stone, small variations are possible and indeed found every day, yet all are equally calid.

The touchstone is whether they work properly or not.

In principle, standard Tube amps have "some but not too much" NFB (Audiophile opnes are in another area) , from as low as 6dB to as high as 12dB.

So it does not *really* matter much what tap is feedback taken from, or actual resistor value, the real point is checking that adding/removing NFB varies gain by 6 to 12dB and if outside that, adjust tap (if many available) or NFB resistor value (much easier) so you get within that range.

That tap used is one or the othger, same as used by a speaker or a dedicated one is not terribly important, all are "a sample of output voltage".
 

PRR

Member
Paid Member
2003-06-12 7:04 pm
Maine USA
www.diyaudio.com
> the "stock" plate resistors of the inverter stage were 470k not the 120K on the schematic

It is a Floating Paraphase. The plan shows 120k plate feed resistors *and* a 470k/510k pair to feed the offside triode. 470k/470k would not be audibly different and may have been expedient in production.

4K plate-plate is real heavy loading for 6L6 at high voltage with self-bias. However the iron is in your hand and in your chassis, you say it works, so let it be.

If we assume the secondary is 0/8/16 then your ohm-meter readings make perfect sense (within the sub-Ohm accuracy of most meters).
 

Attachments

  • Lebow-42.gif
    Lebow-42.gif
    8.5 KB · Views: 17
Thanks, folks. Yes, the NFB seems very forgiving.( Piazza got it right - same as what's pasted on the amp.) The previous owner put in a pot to vary over the 5k-10k feedback resistor range shown on the schematic. Can hear volume increase somewhat but the character of the tone at full volume remains pretty fat and rich. Vibrato as nice as my '59 Tremolux. Haven't looked on scope yet to see what's going on.

The input jacks aren't grounded so there is loud buzz/hum with no guitar cable plugged in. Unused channel needs volume down. Seems like a poor design.
 
Thanks, folks. Yes, the NFB seems very forgiving.( Piazza got it right - same as what's pasted on the amp.) The previous owner put in a pot to vary over the 5k-10k feedback resistor range shown on the schematic. Can hear volume increase somewhat but the character of the tone at full volume remains pretty fat and rich. Vibrato as nice as my '59 Tremolux. Haven't looked on scope yet to see what's going on.

The input jacks aren't grounded so there is loud buzz/hum with no guitar cable plugged in. Unused channel needs volume down. Seems like a poor design.

Easy to solve, just copy Fender input jack grounding :)