U73b (drip) - 3 questions

[Sc]

Hi Axl85

I'm glad you made it that far - you are close to the spec's now.
Cal step 2 - yes, no problem to have more output - it has to be at least +2dBU or more.
In limiter mode you should get quite a bit more more than +7.5dBU - in case you have the correct DC tension of 1.8V at cathode and around 80V to 90V Anode I'd suspect the E88CC valve (or it might be the choke issue). Maybe it's easier to change the valve for a first test. Else you can try to lower the parallel resistor from 47k to 39k or 33k. If this gives you more output, I'd switch the choke to 1377 (or maybe to the original BV514 Haufe - Manfred Deppe told me, that it's still availiable there). I don't know much about the effects of paralleling a plate choke with a fixed resistor - it probably limits some of the benefits.
Let us know - it's a tricky thing there.
best regards Walter

 

[axl85]

Hi Axl85

I'm glad you made it that far - you are close to the spec's now.
Cal step 2 - yes, no problem to have more output - it has to be at least +2dBU or more.
In limiter mode you should get quite a bit more more than +7.5dBU - in case you have the correct DC tension of 1.8V at cathode and around 80V to 90V Anode I'd suspect the E88CC valve (or it might be the choke issue). Maybe it's easier to change the valve for a first test. Else you can try to lower the parallel resistor from 47k to 39k or 33k. If this gives you more output, I'd switch the choke to 1377 (or maybe to the original BV514 Haufe - Manfred Deppe told me, that it's still availiable there). I don't know much about the effects of paralleling a plate choke with a fixed resistor - it probably limits some of the benefits.
Let us know - it's a tricky thing there.
best regards Walter

Hi Walter,
thanks again for the reply! (somhow i deleted my last post by mistake...sorry!)
Great, i will check the tube and get a new pair maybe! I might also write to Manfred Deppe and ask him about the choke. Will report any results of course!
In the mean time another question
Did you find a good solution for grounding the main pcb? I noticed some hum sice i got the new power tansformer in. So i thought i'll check my grounding again. I think the main board has 4 or 6 grounding points. I was never really sure what's the best way to connect them to the chassis... Do you have some recomendations maybe?
Best,
Alex

 

[Sc]

Hi Alex
Uh - ground loops ...  :. If possible - check with an oscilloscope whether you have the ac-noise with no other equipment connected or just hook the output to an amp without ground connection (not over mains ground - be careful - and not over XLR). In case it's in the unit itself, you might try this:
In case you have no choke at the PSU - you might try with a choke around 6H first - might help. 
Or do a star ground. Scrape the color off where one of the power transformer nuts is, and fix a connection bar there. Then fix the mains ground there. Isolate the mainboard from any chassis connection and go from the GND point (next to the small transformers) with a thick ground lead to star ground. Make sure, you have no other ground connections to chassis at the other boards - else isolate and go to star ground the same way.
I suppose the original unit was totally off earth (only the case was) - the way the connectors were set up, would make no sense. And the varicom is as well off ground - the reference is to the zero volt line. So I'll try to set up the full combination including varicom off case and mains ground with a ground lift switch. But maybe this is just crazy  ;D. The parts are underway- so I hope I'll have news in a few weeks.
regards Walter   

 

[andre tchmil]

any idea how a sidechain HP could be integrated in this design ?

 

[Sc]

any idea how a sidechain HP could be integrated in this design ?

Hi

It's sure nice to have a HP in the sidechain - unless you use the varicom, you can switch bass response of the sidechain there. Below the original TAB-schematics. The yellow marked stuff is not on drip's design. The two red circles mark the switches between the 300k resistor and a preemphasis circuit. That's the spot you would want to insert any sidechain filters - the load of the filter should probably be close to 300k to keep calibration to specs. 

 

[andre tchmil]

Hi

It's sure nice to have a HP in the sidechain - unless you use the varicom, you can switch bass response of the sidechain there. Below the original TAB-schematics. The yellow marked stuff is not on drip's design. The two red circles mark the switches between the 300k resistor and a preemphasis circuit. That's the spot you would want to insert any sidechain filters - the load of the filter should probably be close to 300k to keep calibration to specs.

Do I miss something ? That circuit is at the input , not sidechain

 

[mg73]

Yes you do!
Look closely, the input on the lower primary winding is s/c

Do I miss something ? That circuit is at the input , not sidechain

 

[andre tchmil]

Oops , yes of course. Must have been the glühwein.

 

[axl85]

Hi Alex
Uh - ground loops ...  :. If possible - check with an oscilloscope whether you have the ac-noise with no other equipment connected or just hook the output to an amp without ground connection (not over mains ground - be careful - and not over XLR). In case it's in the unit itself, you might try this:
In case you have no choke at the PSU - you might try with a choke around 6H first - might help. 
Or do a star ground. Scrape the color off where one of the power transformer nuts is, and fix a connection bar there. Then fix the mains ground there. Isolate the mainboard from any chassis connection and go from the GND point (next to the small transformers) with a thick ground lead to star ground. Make sure, you have no other ground connections to chassis at the other boards - else isolate and go to star ground the same way.
I suppose the original unit was totally off earth (only the case was) - the way the connectors were set up, would make no sense. And the varicom is as well off ground - the reference is to the zero volt line. So I'll try to set up the full combination including varicom off case and mains ground with a ground lift switch. But maybe this is just crazy  ;D. The parts are underway- so I hope I'll have news in a few weeks.
regards Walter 

Great! Thanks again! I'll check the star ground first and see how far i get.
I just was thinking about R40 again.... I put a 10k resistor in. Did You also skip R40 in your unit? Or just R60? I got a little confused...
I also was not sure about the Jump7 spot on the main PCB. There is a note to use a 10k resistor in case of a 156c plate choke. I'm still using the wrong choke (9375) from sowter. Should i skip the 10k resistor there and make a jumper in?
Have a nice christmas by the way!!
Alex

 

[Sc]

Great! Thanks again! I'll check the star ground first and see how far i get.
I just was thinking about R40 again.... I put a 10k resistor in. Did You also skip R40 in your unit? Or just R60? I got a little confused...
I also was not sure about the Jump7 spot on the main PCB. There is a note to use a 10k resistor in case of a 156c plate choke. I'm still using the wrong choke (9375) from sowter. Should i skip the 10k resistor there and make a jumper in?
Have a nice christmas by the way!!
Alex

Seasons greetings to you ! No, I didn't alter R40 - just R60 - sorry that was my mistake in earlier posts. As long as you have a flat frequency response at the output of the interstage there, no need. But take the other 10k out and jump. With paralleling the platechoke you reduced the DCR - the 10k adds to it again. The 9375 has a measured DCR of about 18k, the 1377 (also measured) 9k and the 156C 3.7k (datasheet) - so only this one needs an additional resistance. This probably improves the output ....   

 

[axl85]

Seasons greetings to you ! No, I didn't alter R40 - just R60 - sorry that was my mistake in earlier posts. As long as you have a flat frequency response at the output of the interstage there, no need. But take the other 10k out and jump. With paralleling the platechoke you reduced the DCR - the 10k adds to it again. The 9375 has a measured DCR of about 18k, the 1377 (also measured) 9k and the 156C 3.7k (datasheet) - so only this one needs an additional resistance. This probably improves the output .... 

Ok great! I’ll try without the 10k!
I did some work on the grounding. Tried star ground and put the mains ground at the same spot as the power TX.
Then I removed 3 of the 4 standoffs around the XLRs and just grounded the remaining one. This is where my star ground is. I also removed the little fixing thing on the XLRs which connects them to the chassis. Now the unit is really quiet! Still some very very low hum but definitely something I can work with. Thanks for the inspiration, Walter.
Else I got 2 new E88C tubes because I didn’t really get 80v at one channel. I got some new JJs but both of them don’t make 80v at the plate. I use a NOS Tesla at the other channel wich makes exactly 80v (1.60v at the cathode). If I put this one into the other channel i get same readings.  Is there such a big difference between tubes...? The two new ones both make about 65v at the plate. Should i get some more and see if one of them work? Or is there any trick to work with?

 

[Sc]

Ok great! I’ll try without the 10k!
I did some work on the grounding. Tried star ground and put the mains ground at the same spot as the power TX.
Then I removed 3 of the 4 standoffs around the XLRs and just grounded the remaining one. This is where my star ground is. I also removed the little fixing thing on the XLRs which connects them to the chassis. Now the unit is really quiet! Still some very very low hum but definitely something I can work with. Thanks for the inspiration, Walter.
Else I got 2 new E88C tubes because I didn’t really get 80v at one channel. I got some new JJs but both of them don’t make 80v at the plate. I use a NOS Tesla at the other channel wich makes exactly 80v (1.60v at the cathode). If I put this one into the other channel i get same readings.  Is there such a big difference between tubes...? The two new ones both make about 65v at the plate. Should i get some more and see if one of them work? Or is there any trick to work with?

Hi Alex
Glad it worked out with the star ground  .
It's two factors - one is anode and cathode resistance. First jump the 10k at the choke - this will alter the readings again. Whenever you should change the choke or alter the parallel resistor, the readings will change also. Opposite to us you have rather low readings - so the cathode resistor could maybe be closer to the original 500R. The second factor ist the valve and unfortunately these have even new very high tolerances (60ties standard: anode current up to 20% +- and transconductance up to 40%+-). Today I think it's even worse. But if both JJ behave the same, use these and try to get a bit closer to 80V and 1.8V (but at this point within 10% it should be fine). You can buy matched pairs or maybe somebody close to you has a valve tester (and hopefully a large box with (N)OS E88CC ....  :).
 

 

[axl85]

Hi Alex
Glad it worked out with the star ground  .
It's two factors - one is anode and cathode resistance. First jump the 10k at the choke - this will alter the readings again. Whenever you should change the choke or alter the parallel resistor, the readings will change also. Opposite to us you have rather low readings - so the cathode resistor could maybe be closer to the original 500R. The second factor ist the valve and unfortunately these have even new very high tolerances (60ties standard: anode current up to 20% +- and transconductance up to 40%+-). Today I think it's even worse. But if both JJ behave the same, use these and try to get a bit closer to 80V and 1.8V (but at this point within 10% it should be fine). You can buy matched pairs or maybe somebody close to you has a valve tester (and hopefully a large box with (N)OS E88CC ....  :).

ok, that makes absolutly sence, walter!
I jumped the 10k at the choke and got some 100v at the plate. so i took the parallel resistor out. Finally i'm really close to 80v and 1.8v.
I'll get some e88cc tubes in the next weeks and see wich are closest to each other. Maybe that gives some better performance in stereo operation. Also have to write to Manfred Deppe about the choke. That's the most critical point i guess.
I use stepped swiches for the theshold wich are almost identical on both sides (used0.01% resistors). That makes both channels easy to compare as well. So at the calibration i realized that one channel comresses more then the other one with same amount of threhold. I guess that's because all my tubes are not matched...?
Anyway... Thanks a lot!!

 

[Sc]

ok, that makes absolutly sence, walter!
I jumped the 10k at the choke and got some 100v at the plate. so i took the parallel resistor out. Finally i'm really close to 80v and 1.8v.
I'll get some e88cc tubes in the next weeks and see wich are closest to each other. Maybe that gives some better performance in stereo operation. Also have to write to Manfred Deppe about the choke. That's the most critical point i guess.
I use stepped swiches for the theshold wich are almost identical on both sides (used0.01% resistors). That makes both channels easy to compare as well. So at the calibration i realized that one channel comresses more then the other one with same amount of threhold. I guess that's because all my tubes are not matched...?
Anyway... Thanks a lot!!

Hi Alex
Wow - good idea with the stepped threshold - gives a clear calibration point too! Differences between the valves is most likely the reason for differences in performance (of course there are some others too). But 5 to 10% deviation (Ia and S) for the same types at the same place in the channels is fine enough - my os valves are about there and the unit does ok. I wrote Haufe yesterday about the BV514 - let you know.

 

[Sc]

Hi Lee and axl85
The swap red/green should work fine - I'll include a pic of the possible connections. Primary and secondary are reversed as well on the board, but I really don't know about possible influence of this on a 1:1 transformer with CT.
When you shunt the 9375 with a 47k Resistor you have a DC-Resistance of about 13k - this is not far from the 1377 (measured). But you might loose some of the "compressorlike" dynamics of a platechoke. Important to have 1.8V DC max. on the cathode resistor there - otherwise one halve wave goes into cutoff at low levels already.
I'm working on integrating varicom and original time constants - let you know, when finished (could take a while .....).

Walter

UUps - the very last connections on the pic are wrong, sorry  :-\. Must be white to brown or green (2 or 5) and 4 blue ...

 

[axl85]

Hi Alex
Wow - good idea with the stepped threshold - gives a clear calibration point too! Differences between the valves is most likely the reason for differences in performance (of course there are some others too). But 5 to 10% deviation (Ia and S) for the same types at the same place in the channels is fine enough - my os valves are about there and the unit does ok. I wrote Haufe yesterday about the BV514 - let you know.

Hi Walter,
any news from Haufe?
I got a mached pair NOS Siemens E88CC and they are super close to the required voltages. Finally Still would be interested in the right choke of course!
There is one thing which came to my attention. If i turn up the theshold with no signal coming in, the VU meter of one channel showes some gain reduction. Do you have an idea how that is possible? I was seaching for some things around the E99fs but didn't find anything suspicious yet...
Happy new year by the way!

 

[Sc]

Hi Walter,
any news from Haufe?
I got a mached pair NOS Siemens E88CC and they are super close to the required voltages. Finally Still would be interested in the right choke of course!
There is one thing which came to my attention. If i turn up the theshold with no signal coming in, the VU meter of one channel showes some gain reduction. Do you have an idea how that is possible? I was seaching for some things around the E99fs but didn't find anything suspicious yet...
Happy new year by the way!

Hi Alex
Happy New Year to you ! Yeah, the good old stuff ...  I'll get an offer from Haufe - they have it, but in a certain version (?). Let you know.  I'm not sure about your threshold issue. Noise floor difference or high frequency oscillation ?? Anyway - it could be the E99F - pair itself - would swap these and see, whether it happens on the other channel too. Good luck ! 

 

[Sc]

Hi Walter,
any news from Haufe?

Hi Alex
Haufe still produces the BV514 on order (9 weeks) not really cheap though. Sent you a pm.

 

[axl85]

Hi Alex
Happy New Year to you ! Yeah, the good old stuff ...  I'll get an offer from Haufe - they have it, but in a certain version (?). Let you know.  I'm not sure about your threshold issue. Noise floor difference or high frequency oscillation ?? Anyway - it could be the E99F - pair itself - would swap these and see, whether it happens on the other channel too. Good luck !

Yes, that's it! I swaped the E99Fs and checked the symetry. Now it's working fine!
Did you made any progress with the ground/lift switch so far? Any good results?

 

[andre tchmil]

I'm playing a bit with an original U73. filters removed on the SC, 300K in circuit.
measurement shows a high shelving boost at 5K the more I lower the threshold, thus more compression.
No idea if thats normal for these units.
How does the Drip version respond ?