Pitfalls to warmup delay in DC coupled multiple railed amps?

[leadbelly]

My question was sort of covered in another thread on B+ warmup delay vs fixed bias warmup, but different enough so that I thought I'd pose the question.

The design I'm working out is a two stage SE amp, driver DC coupled to output stage, both cathode biased. The PS is made up of 3 rails, a bipolar pair for the driver stage with its own tube rectifiers and transformer, and of course a separate transformer and rectifiers for the output stage.

Basically, since warmup time varies with tube rectifier type, and logically I will use different tubes since the current requirements are so different, is there a problem if one stage warms up well before the other one?

 

[DigitalJunkie]

There could be some issues..

The first scenario that springs to my mind is if the output tube/rectifier warms up,and comes up to voltage before the driver tube. Since the first stage isn't drawing current yet,it's plate could be at (or near) the full supply voltage.This could severely over-bias the output tube,since the grid (directly coupled to the plate) will be positive by a fair margin (near the positive supply voltage!).
This is assuming the rectifier warms up in time to cause trouble,If you're lucky the rectifiers might take a bit longer to warm up than the tubes,and would kind of negate this issue.
You might get by with using indirectly heated rectifiers,they should take a bit longer to warm up,and come up to voltage than directly heated rectifiers.

I'm not visualizing any major pitfalls If the driver stage 'comes up' before the output stage...but I could easily be missing something.

Just my tipsy thoughts on the subject.. Back to my drink!

 

[leadbelly]

You might get by with using indirectly heated rectifiers,they should take a bit longer to warm up,and come up to voltage than directly heated rectifiers.

Thanks! I think it's fairly easy to play with warmup times using damper diodes, and combinations of damper diodes with more traditional audio tube rectifiers in the different stages.

 

[richwalters]

Basically, since warmup time varies with tube rectifier type, and logically I will use different tubes since the current requirements are so different, is there a problem if one stage warms up well before the other one?

With power amps, 90% of front ends warm up faster (TV video tubes around 11 secs or faster) than subsequent power amp stages can be 20-30 secs for 6550 series, then a max front end gain instability scenario occurs before o/p stage warms up and global nfb loop comes into effect. This can make squealing noises as the loop gain stability boundary runs out of control and momentarily oscillates.
Some versions of the original 1953 GEC KT88-50 (3 stage) were prone to this, as the B+ roared to 500V, practically instant after switch on. Amps which squeal on warm up or power down, indicate a touchy stability issue. Often too much global feedback to tidy up other performance parameters.
The ultimate warmup solution is per Morgan Jones semi digital technique or in my case allow a B+ lockout watchdog " 15-20 secs wait state".
In my area with so many line brown-outs, this is vital as to save persistent fuse blowing with high VA power transformers.

In most power amp designs heater wiring is often miscalculated and with poorly dimensioned Cu wire to heater wiring can often starve vital heater power to cold power tubes. Designing multiple o/p stages, one immediately realises 6.3V is a downfall lousy heater voltage.


richy