Motown Direct Amplifier-inspired Preamp?

[DaveP]

The first trace looks better than the second I think?
best
DaveP

 

[map2112]

Unless we hear from CJ otherwise, this is the final schematic:-

best
DaveP

The closed circuit 1/4” output jack will effectively short out the secondary of the output transformer if nothing is plugged into it (if wired as shown).

Perhaps you intended to have pin 2 of the XLR connected to the switched connection and not pin 3?

Mark

 

[dai h.]

I think that is a stereo jack (tip and sleeve--ring not shown) and not a switched mono jack.

 

[MidnightArrakis]

I think that is a stereo jack (tip and sleeve--ring not shown) and not a switched mono jack.

Unless the schematic is "properly" drawn out to show what the designer's actual intent is, you can only -- surmise -- what the current schematic is actually showing. DaveP will either need to update his schematic to electrically show what his intent really is or, he will just need to offer an explanation of what his current schematic is actually conveying. But.....I also get your point as well!!!

/

 

[DaveP]

I'm glad someone is looking at the detail, thanks 2112! I will correct that in the next version.
best
DaveP

 

[DaveP]

We are now up to mk9 and getting better all the time, thanks to all contributors.

best
DaveP

 

[CJ]

did some tests with a pot in place of the feedback resistor, here are some numbers>

Output Voltage taken at output cap, not xfmr secondary, voltages are P-P values

Input: 100 mv 1 KC

Output: 10 V Gain: 100 F/B Resistance: 240K
Output: 5 V Gain: 50 F/B Resistance: 91K
Output: 3.6 V Gain: 36 F/B Resistance 56K
Output: 1.0 V Gain: 10 F/B Resistance 14K

So due to some changes in grid resistors and the tube type, the F/B resistor value for a gain of 50 (34 db) is now 91K, at least on the amp that i built, this includes the 20 pf cap across the resistor.

also, with the new grid resistors, the sq wave response seems to have gotten better to the point to where the grid stopper on stage 2 is no longer needed.

i am going to use a pot to adjust the gain of this circuit from 50 to 120 so it matches the vol pot on my other amp.

 

[CJ]

Here are some numbers for a clipping test with various levels of negative feedback, volts P-P, 1 KC Sine Wave>


NFB Resistor: 240K Clipping at 1.6 V Input Output Voltage at cap: 120 V Gain: 75
NFB Resistor: 91k Clipping at 2.7 V Input Output Voltage at cap: 110 V Gain: 40
NFB Resistor: 56K Clipping at 3.5 V Input Output Voltage at cap: 100 V Gain: 28.5
NFB Resistor: 15K Clipping at 5.0 V Input Output Voltage at cap: 50 V Gain: 10

so gain drops at clipping level except when the resistor is at 15K which is when the most feedback voltage is applied to stage 1.

 

[DaveP]

This has been a very educational thread, CJ has shown how you go about developing a circuit and curing its snags on the bench. Over the years I have been on this forum, it has been disappointing to see so many guys wanting easy answers instead of working things out for themselves. Solving problems by experiment gives you confidence and cuts through the BS that you find everywhere online nowadays. Groupdiy is a great source of ideas that can help you solve the problems yourself, all you have to do is listen.......
best
DaveP

 

[s wave]

I wholeheartedly agree Dave P. I learn so much from a thread like this and the amazing 'roll up your sleeves' circuit pioneers, experimenters and engineers!!

 

[aaquilato]

Agreed. This has been an eye opening thread. I don’t understand everything as I have no EE trading but CJ, thank you! Ian, thank you!
I might actually try to build one of these.

 

[CJ]

Thanks, I did have a lot if help from other people, and due to the suggestion of using a smaller grid resistor, I found a bad solder joint on a 1 meg on grid #2, the resistor was loose in the socket pin, I am wondering if this was the cause of the excessive ringing at 10KC, I might unsolder that resistor to see what the effects are.

Funny that I made some changes and then undid them which led back to pretty much the original schematic,


Check this out, they make Motown stickers, I got one coming in from Redbubble for the front, put a little clear coat on it,

https://www.redbubble.com/i/sticker...p8I7xe0zDYEMrZ5vjbkaAidMEALw_wcB&gclsrc=aw.ds

 

[heatwalk]

Are you getting frequency response shifts with the NFB pot changes? The SA-70 design with the NFB control tweaks gets some variations happening. But with that large of a coupling cap am I correct in assuming the low end is fairly constant? The paralleled pF cap and the pot I would think would change HF response.

This brings up a question I have regarding where NFB is sent. I have seen designs such as this one with NFB going to the cathode but in the aforementioned SA-70 it is brought back to the previous plate - can anyone chime in with theory differences here?

I have also been pondering NFB adjustment in circuits like this using a pot in lieu of (or in series with) a cathode resistor as an alternative implementation to increasing resistance in the path back. Eg sending the loop closer to ground.

 

[CJ]

you could do it like that although you would have all the tube current going through the pot, which isn't a lot, i might be worried about a scratchy pot, and you would be a bit constricted in your pot choice, 100, 500, 1K ohm etc

i used a 100 k resistor in series with a 250K pot that really measure 240K, it was pretty easy to dial in the span i wanted, it goes from about 0 dBu to minus 8 dbu which matches up with the vol control on my other circuit,


this way of controlling gain is a lot better than putting a pad on the front end as that would load down the guitar or bass which would mess with the tone,

i will do some scope shots for freq response and sq wave testing with various gain settings tomorrow.

 

[CJ]

Listening tests at various gain settings say to keep 34 dbu as the midpoint and work up and down from there. My stock Telecaster has no problem hitting 0 dbu at the lowest gain setting so I really don't need more but a hot pickup through a stomp box might need less gain so I am going with 34 dbu with plus and minus 6 dbu for the gain pot although thus project is fine with fixed gain as any need gain adjustments could he made at the mixing board.

Stock 34 dbu gain seem to work well with acoustic guitars, higher gain not so much. Acoustic guitar drove the meter at around the minus 6 to minus 3 dbu level at the stock gain setting.

Dave's circuit sounds a bit better than mine for acoustic with the better defined midrange and top end.

There is an increase in brightness ad the gain goes up but this might be useful in some situations.

Next am going to try a 56k fb resistor with a 100k pot on the end.

 

[DaveP]

Are you getting frequency response shifts with the NFB pot changes? The SA-70 design with the NFB control tweaks gets some variations happening. But with that large of a coupling cap am I correct in assuming the low end is fairly constant? The paralleled pF cap and the pot I would think would change HF response.

This brings up a question I have regarding where NFB is sent. I have seen designs such as this one with NFB going to the cathode but in the aforementioned SA-70 it is brought back to the previous plate - can anyone chime in with theory differences here?

I have also been pondering NFB adjustment in circuits like this using a pot in lieu of (or in series with) a cathode resistor as an alternative implementation to increasing resistance in the path back. Eg sending the loop closer to ground.

The 3uF feedback cap C3, was chosen for a feedback resistance of 54k, but is still large enough for lower resistances that you would not notice the difference at the low end. The 20pf gives a -3dB at 150kHz, so again it would be difficult to detect frequency response changes, maybe dogs could?

If the NFB goes from plate to previous plate, it is negative feedback, with the feedback resistor acting on the tube resistance, not very consistent as the tube resistance can vary along the load line. If it goes from plate to previous cathode it is positive feedback which acts like negative as it makes the grid more negative relative to the cathode. See your annotated sketch.

Have a look at the V series pre=amps, they use the technique you describe (the V78 has a pot) It is better to use switched resistors because pots can be scratchy.
best
DaveP

 

[CJ]

Optional gain mod:

i plugged in an MXR Distortion + at full volume and found that it puts out about 330 mv RMS with a Hamer 335 copy strummed at maximum volume on an E chord,

so I scaled the amp to take this signal and put out 0 dBu on the VU meter which is sitting on the output transformer secondary.

27K resistor and a 250KA (measures 240K) pot did the trick,

i got a guy in the Stone Jack Ballers Grateful Dead cover band with a hearing problem so he keeps turning his amp up, so the reduced gain is waiting for him, and he uses JBL's with the aluminum domes, can you say icepick?

when the pot is turned all the way up for full gain it only takes a 50 mv RMS signal to drive the output to 0dBu which is good for low output instruments such as acoustic guitar pickups. (non active)

next a look at the effects this gain mod might have on freq plots and sq waves

gain mod schemo and the pimped out box that is waiting for the Hitsville decal>

and is mis-pelled Motown. sheesh, oh well, at least i don't have to worry about Berry Gordy coming for me.
i'll tell him it is a box for jump starting lawnmowers,

 

[CJ]

10 KC sq wave at three different gain settings>

input on top, output from cap, not transformer, on bottom,

input level set to give an output of 0 dBu in all cases

note that the transformer primary was hooked up so this could have an influence on the waveforms,

i might try disconnecting it to see,

 

[CJ]

100 KC sq wave at three different gain settings>

 

[CJ]

10KC sine wave phase shift at 3 different gain settings>