• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Optimal Umbilical Runs and Connections

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Not sure if the subject is actually useful here -- I am running an umbilical connection from a PS to an amp. The PS will send a raw DC voltage to the amp where it will be regulated.

The tubes in the amp will be heated with AC. Typically, I build a voltage divider from B+ and connect the heater winding's center tap to bias the heaters up about 40V or so, which seems to work well. So, the question is, should I do the biasing in the PS (which is cLCLC filtered so reasonably quiet on its own, and reasonably well regulated due to the choke input), or is it better to run the center tap all the way to the amp to build the divider off the well regulated B+? Does it even matter?

Finally, any problem with 6.3VAC and ~200VDC in the same ~1m connection?
 
There are connectors like Speakons with shielded pins, and on the other hand there are XLR types with only one shielded end of a normal cable. If you have any B+ capacitors in the circuit box you'll need a shielded connector there. Obviously has to be shielded at the PSU box.

You probably know this, but just to make sure!!

200vDC allows you to use all kinds of connectors - XLRs, D-connectors, speakon 4 or 8 way etc. Most are specced for 250vDC. Seems to be just a beurocratic decision to state 250v max. since originally these connectors were tested for much higher DC.
 
I would use Belden 1037A cable, individually shielded twisted pair, good for 600V.

No problem running DC and AC in the same cable. However, I would send 120VAC into the cable, instead of 6.3. Place your filament xfmr inside the amp chassis, not the PS chassis.

This, then answers your last question about how to bias the 6.3V winding.
 
You didn't mention if you wanted the ability to unplug the umbilical from the amp and/or power supply but I assume you do. There are lots of choices of connectors to do this from Molex to the Speakon connectors mentioned before. Whichever connector you choose it must of course be keyed. Wouldn't want to plug 200VDC into the wrong thing! :hot:

That 200VDC is important in another way. It can really hurt if gets out! Therefore the umbilical should be designed to prevent the possibility of anyone touching a hot pin. My solution it to permanently connect the umbilical to the amp and only have a plug on the PS end. Make the jack on the PS have the **female** connector, no chance of anyone touching a hot pin there. Have the connector coming from the amp be the male end.

I've put AC and DC in the same umbilical successfully. I used shielded twisted pair for the AC and regular teflon insulated 600V non-shielded wire for the DC. It worked fine though if I were to do it again I would do as zigzagflux recommends and run 120VAC through the umbilical rather than high current 6.3VAC.

I've used Molex connectors for this type of thing and they work well but they aren't very attractive. Speakon connectors look much more professional. In lower voltage (but much higher current) chip amps I've even used 5 pin DIN connectors with the female jack on the PS. The wires are thin but with runs of < 1 meter they have worked flawlessly.
 
I have the same exact questions for a Raven-style Line stage preamp that I'm about to implement.

Since 2 of you have suggested running 120VAC through the umbilical (instead of 6.3 VAC) I'll probably go with your suggestions. I'd like to understand why you make these suggestions, however.... (voltage/current drops?)


Additional questions:

I'd need to send B+ (around 125VDC), -12.5 VDC (for cathode CCS bias), ground, and 120 VAC through the umbilical. Does the shield of the umbilical cable become the ground signal? Or should there be a separate wire in the umbilical that is used as the ground signal?

If I'm using a star grounding scheme, do I have all the grounds come together in the PSU box?
Or are there 2 separate star grounds (one in the PSU box and one in the amp box) that are connected together via the ground in the umbilical?

Chris
 
Administrator
Joined 2004
Paid Member
I use and recommend CPC connectors for these applications. They are keyed, fairly inexpensive and can handle high voltages and currents.

I like to keep all supplies including filaments in the main supply, and run those to the device being powered. Tightly twist the filament wires and run with everything else in the umbilical. I've been doing this forever and have never had a problem with coupling between supplies.

It would not be a bad idea to interlock the connector so that if it is not connected the power supply cannot operate.
 
zigzagflux said:
No problem running DC and AC in the same cable. However, I would send 120VAC into the cable, instead of 6.3. Place your filament xfmr inside the amp chassis, not the PS chassis.

The whole point of the separate PSU chassis is that I am getting magnetic coupling between the power and output transformer, so I can't put the filament transformer in the same box.

ckniker said:
I have the same exact questions for a Raven-style Line stage preamp that I'm about to implement.

You won't have the same magnetic coupling issues because it is a push-pull design, so no (or at least less) worries about transformers being in the same box.

ckniker said:
Since 2 of you have suggested running 120VAC through the umbilical (instead of 6.3 VAC) I'll probably go with your suggestions. I'd like to understand why you make these suggestions, however.... (voltage/current drops?)

It's because of the lower current that needs to travel through the wire.
 
kevinkr said:
It would not be a bad idea to interlock the connector so that if it is not connected the power supply cannot operate.

So, like a loopback from one otherwise unused pin to another otherwise unused? Doesn't the pin still have high voltage on it, or do you use a relay of some sort?

I am planning to use some 5 pin XLR's here (yes yes, SY, I know all the warnings, but the number of people who will touch this amp is exactly 1 -- it sits in my home office) and I'll use a female jack on the PS and male on the amp for safety.
 
Great ideas guys! I learned a few things as usual.

I especially like the loopback idea. I was thinking instead of having all the mains current travel through the connector cable, finding a 120V controlled relay and run it in parrallel with the mains which would be switched by relay. The current traveling through the cable would just be the current needed to drive the relay. Are there such safe relays?
 
I go to Industrial Electric Wire and Cable, http://www.iewc.com/
Last time I checked, the cable was in stock. Minimum order applies, but it wasn't all that much. You will definitely end up with more than 2 meters.

My suggestion on the shields is they are strictly used for shielding and grounding of the chasses, not used for audio signal conductors. Twisted pair plus shield should be all you require to keep each signal hot and cold in its own shield, happily isolated from the rest of the world. Circuit common, however you choose to implement, will use the cold conductor, not the shield. See http://www.nutshellhifi.com/Karna.gif for visual.

You never specifically stated if your tubes were direct heated or not, but if they are direct, keeping your filament transformers inside the amp proper makes for a shortest path or audio . Even if indirect heated, how small of a chassis are you planning that you can't avoid magnetic coupling to the output transformer? Push-pull or not, the issue doesn't change. Additionally, the more filaments you have to heat, the potentially more conductors you need in the umbilical; one 120VAC is all you need.
 
zigzagflux said:
Even if indirect heated, how small of a chassis are you planning that you can't avoid magnetic coupling to the output transformer? * * * Additionally, the more filaments you have to heat, the potentially more conductors you need in the umbilical; one 120VAC is all you need.

2 tubes (7788's) so current draw is under 1A, and while it isn't that small, it is a headphone amp, so any hum is way too much.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.