Api 312 Brent averill edition..

[pucho812]

This is weird. With -20dBu test signal I measure a max of 1dBu output.  This is with gain pot at full. I should see some 29-30dBu at the output under normal working conditions.

So far caps are good, diodes are good and swapping the 2520 yielded no change.

I have a second  Brent Averill on the bench that is  working and the only different between the two is input transformers, one has a Jensen the other has an api input transformer.

Any ideas.

 

[Rob Flinn]

Have you looked up the spec of the 2 input transformers to see if they are the same ratio ?  Possibly not.

 

[pucho812]

Have you looked up the spec of the 2 input transformers to see if they are the same ratio ?  Possibly not.

It would be very odd for a company that made both units to use transformers that were different ratios, me thinks. But I'll take some measurements and see what is what. Ideally if the ratios are the same on the transformers, I should see similar measurements at their outputs.

 

[gridcurrent]

It would be very odd for a company that made both units to use transformers that were different ratios, me thinks. But I'll take some measurements and see what is what. Ideally if the ratios are the same on the transformers, I should see similar measurements at their outputs.

the 2622 is a 1:8. 
recall the drama here a while back as to what ratio it really is.
Brent ran out of the genuine API parts and created preamps with the  1:10  115ke coil,  retaining the 312 label.
if I install a Kia engine in my AC Cobra is it still a Cobra ?

 

[pucho812]

the 2622 is a 1:8. 
recall the drama here a while back as to what ratio it really is.
Brent ran out of the genuine API parts and created preamps with the  1:10  115ke coil,  retaining the 312 label.
if I install a Kia engine in my AC Cobra is it still a Cobra ?

right, a 1:10 will add 20dB of signal, I forget what a 1:8 setup does. I remember a 1:5 adds about 14dB. so I am going to guess 1:8 adds about 18dB.

So while I don't expect a +30dBu at full  like on the Jensen, I should see something close if the  second unit was functioning properly.  although I tried another  do a it may have been faulty. few more tests and signal traces to see what is happening.

 

[fazer]

Sounds like a problem transformer either input or output

 

[pucho812]

Sounds like a problem transformer either input or output

that is what I am afraid of.... 

 

[pucho812]

todays testing start with measuring at the secondaries of the funky unit.

the good unit has a Jensen trafo in there, At the secondaries with my -20dBu sine wave of 1K, I measure about 0dBu. This would line up with 1:10 transformer and doing a 20 dB of signal gain

the funky unit has the API input transformer, measuring the secondaries  I measure -2dBu on. the output. This lines up with the  1:8 transformer ratio and an expected  gain of 18dB....

Now the only thing between the  input transformer  and the opamp is a 1/4" jack that has a normal.

hmmmmm....

 

[pucho812]

and we got it....

Boy it's always the simple things. so the funky one which was not working, I bypassed the front 1/4" instrument input and boom it's working as expected. output level of some 30dBu at full gain.  So its isolated to the switchcraft jack and sure enough, the jack which has a normal for the tip is not normalizing.  sure enough the jack normal is physically not making contact.  so time to order a new switchcraft 13B but in the meantime, I added a dab of solder to bridge the gap  and the jack still works.

it's always the last place you look and usually the simple things you over look that get you.

 

[john12ax7]

Normal jacks seem to have a habit of doing that,  they are useful but often problematic.

Curious do you hear much of a tonal change with the 2622 vs 115k? Think the new api also uses 1:10.

 

[pucho812]

Normal jacks seem to have a habit of doing that,  they are useful but often problematic.

Curious do you hear much of a tonal change with the 2622 vs 115k? Think the new api also uses 1:10.

i can run some sweeps with the AP if you like?
But for me there is subtle tonal differences, can't say I prefer one or the other.

 

[pucho812]

so here is a basic AP sweep from 20hZ - 80K

gains were matched with in less then .25dB

the blue sweep is  with the 312 with the API2622 input transformer, the brown/red sweep is with  the Jensen input transformer....

 

[john12ax7]

Thanks,  looks like maybe the zobel are not optimized.

Any chance of distortion plots? Maybe harmonics at 100 Hz and 1 kHz.

Jensen also makes a 1:8, so it's interesting they went with the 1:10, which is a bit high for typical bjt noise performance.

Do you have a subjective preference to either? I remember years ago comparing 512c to the 312, and preferred the 512c, at least on electric guitars .

 

[pucho812]

Thanks,  looks like maybe the zobel are not optimized.

Any chance of distortion plots? Maybe harmonics at 100 Hz and 1 kHz.

Jensen also makes a 1:8, so it's interesting they went with the 1:10, which is a bit high for typical bjt noise performance.

Do you have a subjective preference to either? I remember years ago comparing 512c to the 312, and preferred the 512c, at least on electric guitars .

I have no preference other then they are fixed and go back to my buddy's studio when he picks them up. He has not said one is better then the other so I assume they have no preference either.

no distortion plots but can easily take them as I still have them. I am awaiting for a new switchcraft jack.

 

[drpat]

Common problem with the BAE re-racks (Neve re-racks also).

I clean the normalling contacts in the 1/4" jacks by saturating a thin piece of cardboard (like the back of a legal pad) with EML200F. Rub it between the contacts, and you're good to go for another 5 years.

 

[pucho812]

Common problem with the BAE re-racks (Neve re-racks also).

I clean the normalling contacts in the 1/4" jacks by saturating a thin piece of cardboard (like the back of a legal pad) with EML200F. Rub it between the contacts, and you're good to go for another 5 years.

was easier to replace the jack , and the client didn't mind paying for it. 

 

[JW]

I guess if using a Jensen JE115K, it's a 150K termination resistor? And we remove the 1000pF caps? Any other changes at the 312 input? https://www.jensen-transformers.com/wp-content/uploads/2014/08/jt-115K-e1.pdf

 

[MisterCMRR]

Notice the difference in the shapes of the HF roll-off between the Jensen and API transformers. The Jensen has no peak and roll-off follows a 2nd order Bessel response - which has near-perfect time-domain response (observable as square waves with no overshoot or ringing). The API does not - and will have high deviation from linear-phase (or "time smear" as it's sometimes called). The design of the Jensen, as well as the specification of a 150 kΩ load resistor, is responsible for this intentional Bessel roll-off shape. On really fine material, the difference is audible.

 

[Flatpicker]

I can certainly corroborate MisterCMRR's statements regarding Jensen vs. 'vintage' API clones. I've conducted extensive listening test between the JT-110K and a couple other manufacture's API clones of 1:7/1:8 ratios using acoustic guitar and vocals, and the 'time smear' of the API clones is more than obvious when soloing between tracks of identical recordings. The JT-110K retains all of that nice up-front sound (maybe even a bit more so?) without the mud. Of course the Jensen's are roughly 2X the cost, but chalk it up to another case of getting what you pay for. Now, if only Jensen would make them with split primaries so they could be wired 1:4...

Edit: ...and yes, I used the recommended termination and zobel circuit for each transformer.

 

[JW]

I'm not seeing a recommended zobel in the Jensen document above, just the termination resistor. The 2622 needs the zobel, but Jensen doesn't correct? I refer to the 115K.