REDDI DI

[bubbi]

Magmet don't have the M6 steel you want for this and they usually don't do very small orders.

Why do we want M6 grain oriented steel instead of the High Nickel alloys?

Because high nickel is a lot more expensive and for this kind of project the difference isn't too big at all. This is a question for CJ though !

 

[peterc]

Dirty1

Small point. On the schem, your pot is wired incorrectly.

The link between the 2 lugs should be centre lug to the right hand lug.

Peter

 

[dirty1_1garry]

Dirty1

Small point. On the schem, your pot is wired incorrectly.

The link between the 2 lugs should be centre lug to the right hand lug.

Peter

Oh, yes! Checked twice. 2 lugs should be centre lug to the right hand lug. My fault

 

[abbey road d enfer]

Dirty1

Small point. On the schem, your pot is wired incorrectly.

The link between the 2 lugs should be centre lug to the right hand lug.

Peter

Oh, yes! Checked twice. 2 lugs should be centre lug to the right hand lug. My fault

Not that it really matters. Just changes the rotation direction of the pot. The interesting thing there is that level control is done by loading the output.  Even with large input signals, there is no risk of clipping due to excessive output voltage; current is the only limiting factor; running each section at 8mA ensures nearly 4Vp-p input capability.

 

[peterc]

I understand it would change the rotation direction, but it is a log pot, so the 'feel' would change as well.

 

[bernbrue]

Hi,
thanks for tracing the circuit. Is there any commonly available output transformer that would fit here? What are the specs, ratio etc? I see a cinemag transformer in your pictures. It starts to get very interesting. Keep on!
regards
Bernd

 

[dandeurloo]

The one made by cinemag is exclusive so we would need to come up with something different.

 

[Hank Dussen]

According to CJ we should be looking for a gapped transformer with a 2:1 ratio.
Something like 2k4:600 or 2k:500.
And it should be big enough for a good bass response. Ideal something like 80 henries : 20 henries
(I'm just quoting CJ here)

The Carnhill VTB 2290 seems to be a candidate: http://www.audiomaintenance.com/acatalog/CA-18-VTB2290_extended_info.html
With the primaries in parallel.
No information about the henries tough.

 

[dmp]

the measurements by dirty1_1garry, it looks like a 4:1 transformer.

For understanding of levels inside of schematic I've done some mesurements. Could contain errors but here is what I measured:
send 1kHz 0.775V signal to the input via mono jack
get 6.76V - 8.66V between xlr out pin2 and pin1 (1k log pot min - max)
get 6.28V - 9.8V between out transformer primary wire and gnd.
get 12.8V - 24V between out transformer primary wire and B+.
get 2.6V - 5.3V between out transformer secondary wires.

Primary = 24v-12.8v = 11.2
Secondary = 5.3v-2.6v  = 2.7
11.2/2.7 = ~4

 

[dmp]

The CM-9589 looks like it might work???.
"Vacuum tube output transformer butt  stack optional (takes d.c.) for single ended valve output stages."

cinemag.biz/output/PDF/CM-9589.pdf

 

[CJ]

CM-9589 is for cap feed, so no DC = no gap, plus Ni will not take AC and DC,

99366 listed as butt stacked = DC but no ratio shown at CineMag site

Larry was taking measurements with the pot on and the pot off, thus the v-v readings,

questions on data: flickr pics show 114 VDC on pri to gnd, should be lower,

i get 172 minus 16.35 VDC = 156 (first data post) on the tube plates/pri wire,

172 minus 114 = 58 volts DC across pri, if tubes draw 16 ma, DCR would be 58/.016= 3600 ohms,

that is too high, i like 16.35/.016= 1 k ohms better.

if i use data from Larry's first measurements, i get 9.8 vac/5.3 vac = 1.85:1 turns ratio

this would be 3.42:1 Z ratio, so 3.42 * 600 = 2 k pri Z.

if i use the flickr ac volts on sec, i get 9.8 vac/3.058 vac = 3.2 turns ratio

this = 10:1 z ratio , 10 * 600 = 6 k pri z,

quite a difference from 2 k to 10 k, could Larry please do the pri AC and sec AC measurements again? 

also measure the xfmr core height and width, from the pics of the xfmr next to the tube i bet on 87 EI (7/8" tongue, 2 & 5/8 inch by 2 & 3/16 inch) maybe 75 EI which would be smaller,

if 20 henries on the sec, that would make 68 henries on the pri if we use the 1.85:1 ratio, bandwidth specs would seem to indicate a lower ratio,

5 volts out is about 16 db, which is what they claim, so my $$$ is on the 1.85:1 ratio,

if Larry could break the pri xfmr wire and insert his AC amp meter and pump in a 1 k hz tone, he could measure AC current, then we would have turns ratio and inductance, measure the AC volts across the pri also, XL @ 1 k hz  is what we are after, XL = VAC pri/I-AC pri,

if we have the lams figured then we know the turns count to get that inductance, with the gap of course,

that is steel core, a gap reduces perm of Ni and steel down to about the same figure, 200-300 range, so no use in using Ni plus steel takes 18 k gauss and Ni only 5 k gauss, so steel is the lam of choice here.

strange that they use 2 cathode resistors with the cathodes wired in parallel, but the amp factor of 5.3 vac/.775 = 6.8 seems right for 1 tube stage, not 2 in series,

i can wind this guy on a stack of 75 EI and try it out with a 6DJ8, then offer it to interested DIY'ers, jus sayin...

 

[Hank Dussen]

i can wind this guy on a stack of 75 EI and try it out with a 6DJ8, then offer it to interested DIY'ers, jus sayin...

I'd sure be interested!

 

[ruffrecords]

From the posted measurements the maximum input to output gain is about 21dB. The tube has a mu of 35 so the maximum possible gain from the tube is about 31dB. In practice you get a lot less than this with a transformer load and I would be surprised if you got even 27dB. This means the loss in the transformer can be no more than 6dB implying a ratio of no more more than 2:1.

I am also puzzled by the pot wiring. It seems to me that as drawn this must short out the transformer secondary at one extreme so the output would be zero, but the measurement indicate it is not.

Cheers

Ian

 

[abbey road d enfer]

I am also puzzled by the pot wiring. It seems to me that as drawn this must short out the transformer secondary at one extreme so the output would be zero, but the measurement indicate it is not.

Indeed, measurements are not very reliable.
"get 6.28V - 9.8V between out transformer primary wire and gnd.
get 12.8V - 24V between out transformer primary wire and B+."
These should be identical. (AC-wise, B+ is 0v). Maybe DC component plays tricks with the detector???

"get 6.76V - 8.66V between xlr out pin2 and pin1 (1k log pot min - max)"
This is just a parasitic voltage, of no particular interest.

"get 2.6V - 5.3V between out transformer secondary wires."
This is the significant measurement, but results cannot be considered reliable.

 

[ruffrecords]

I am also puzzled by the pot wiring. It seems to me that as drawn this must short out the transformer secondary at one extreme so the output would be zero, but the measurement indicate it is not.

Indeed, measurements are not very reliable.
"get 6.28V - 9.8V between out transformer primary wire and gnd.
get 12.8V - 24V between out transformer primary wire and B+."
These should be identical. (AC-wise, B+ is 0v). Maybe DC component plays tricks with the detector???

"get 6.76V - 8.66V between xlr out pin2 and pin1 (1k log pot min - max)"
This is just a parasitic voltage, of no particular interest.

"get 2.6V - 5.3V between out transformer secondary wires."
This is the significant measurement, but results cannot be considered reliable.

Understood. I misread the XLR pins - thought you were measuring across 2 and 3. I still don't understand how you manage to get a non zero value across the transformer secondary with the pot shorting it.

5.3V output implies a gain of just under 17dB. 23dB gain from the tube seems a more likely value leading to 6dB loss in the transformer and a 2:1 ratio again.

The pot still does not make sense to me.

Cheers

Ian

 

[dirty1_1garry]

Here is some dimensions of transformer.

Will post some new measurements soon.

 

[abbey road d enfer]

Understood. I misread the XLR pins - thought you were measuring across 2 and 3. I still don't understand how you manage to get a non zero value across the transformer secondary with the pot shorting it.

Neither do I. I think our friend has to do some homework in order to produce more credible peasurements.

The pot still does not make sense to me.

What doesn't make sense? The measurements? - I agree. The way it is connected? - as I said earlier, considering the stage acts more as a current source than voltage, it's a nice way to optimize headroom and THD.

 

[dirty1_1garry]

Done some new measurements. Check it out on flickr.
http://www.flickr.com/photos/95158972@N07/

0.775V in
169 VDC - gnd/b+
154 VDC - gnd/pry wire
2.25 VAC - b+/pry wire min pot
13 VAC - b+/pry wire max pot
0.015 VAC - secondary min pot (connected to the circuit)
2.9 VAC - secondary max pot (connected to the circuit)
4.25 VAC - secondary (disconnected from the circuit)
15.6 VAC - b+/pry wire ( secondary disconnected from the circuit)
Voltage on the 220R = 1.84V DC

 

[dirty1_1garry]

if Larry could break the pri xfmr wire and insert his AC amp meter and pump in a 1 k hz tone, he could measure AC current, then we would have turns ratio and inductance, measure the AC volts across the pri also, XL @ 1 k hz  is what we are after, XL = VAC pri/I-AC pri,

Describe slightly more this method. Where pump a signal?

Tomorrow I should return DI to my friend, so I have one more day for experiments
So if you have any ideas about, let me know))

 

[CJ]

thanks Larry!

put your meter on ohms and measure the resistance on the pri and sec of the transformer,

no power of course,

pri- measure xfmr wires from B+ to plates of tube

sec- measure across the pot with the pot turned to max ohms,
we can deduct the parallel resistance so you do not need to unsolder any wires

that should be enough data to get real close to this thing,

that transformer uses 100 EI, one size bigger than 87 EI, pretty hefty for a line out, means good bass,

just to be sure, if you can, look at the pic and measure where the yellow lines are,

turns ratio looks like 13 vac/ 4.25 vac = 3:1, so 9:1 Z ratio = about 5 k pri Z,

no need to measure current in pri, we have enough data,

thanks again for all the work!