This is quite normal for a tube circuit that is warming up. As emission increases the DC voltages are settling to their optimum points and the charge across the output cap is constantly changing while this happens. You can ruin your speakers if the cone excursion is bad enough and happens enough times. Always turn the power amp on AFTER about a minute of line stage warm up. Or build and include a delay relay that shorts the output for about a minute and you won't have the problem. I prefer a relay that shorts the OP and releases after a min. rather than connects the OP as the latter adds another distortion creating junction.
Mark
BGcap said:
In a common-cathode preamp like the 12B4A, there are two audio capacitors that need to be charged up initially, the big cathode cap, and the smaller output coupling cap. As they charge, voltages (transients) will be present in the output. The charging of these caps depends on the voltage ramp-up characteristics of the B+ and heater power supplies. Various power supply implementations will have different warm-up transient characteristics, but there will almost always be some sort of turn-on transient.
If a tube power amp follows the preamp, its slow warm-up (and infra-bass roll-off) may prevent any objectionable turn-on transient going to the speakers. If you are having problems with a solid-state power amp, you could implement a shorting output mute on the preamp while it is warming up. This can be a simple manual muting switch on the front panel, or a relay actuated by a time delay circuit. In either case the output is shorted until the 12B4 plate voltage fully stabilizes.
In your case, you mention that you held off the power-up of the SS power amp for a minute or longer, but then you still experience woofer movements. It’s possible to have a leaky output coupling cap, although this would be unlikely to occur in both channels. To check for this, disconnect the power amp, let the preamp fully warm up, and measure the preamp’s outputs for the presence of any DC. You should read nothing (just a bit of mV-level noise perhaps). I suggested a 1MEG ground reference resistor in the output. This, in conjunction with a 4uF output cap, minimally loads the plate, but makes for a very long time-constant warm up. You might make this resistor 100K to shorten the warm-up transient time and see if this helps (or reduce the coupling cap size).
Finally, you mentioned seeing woofer movement long after warm-up. Since you are using an unregulated B+ supply, you might be seeing line voltage variations such as when your refrigerator, water pump or oven kicks in (or out). The gas tube regulator shown in the link just above should help. I don’t know if you are using a regulated heater supply or not. I use a supply based on the trusty LM317 (again). I reference this supply slightly positive relative to the cathode (maybe at +20 volts above ground on the negative end of the heater supply, but it’s not critical). The 12B4A is purported to have problems with heater-cathode insulation, although those problems haven’t manifested in my experience with this tube. Keeping the regulated heater voltage close to, but slightly positive to, the cathode will help. With regulated B+ and heater supplies, you really shouldn’t see any transients, other than musical, from line voltage variations.
You could be experiencing low-frequency “1/fn” noise, sometimes called “flicker” noise, due to manufacturing variations among tubes. Try replacing the 12B4As or swapping between channels. You might have to select for best noise performance. Could it be microphonics? Perhaps.
Some combinations of speakers and power amps are more susceptible to transients. Many solid-state amps have sub-hertz roll-offs. Woofers with very low bass capability, especially ported woofers which are undamped below cut-off, can exhibit wild movements in response to very low frequency inputs.
(I see, after writing this, that Mark has offered similar good advice.)
Mark, Brian, Zen Mod, Thanks for the advice.
Waiting a bit longer before firing up the SS amp does help. The woofer still moves, but not enough to be concerned about. The "minute" delay I mentioned earlier was just my perception. I would wait until the tubes were glowing then just a little longer. Obviously this wasn't enough time.
Brian,
I would like to build a gas tube regulator as you suggest, but I may have one issue with this. The tranny (Hammond 261G6) I have is rated for 130ma so would be really close to its max current rating. Would it be preferable or necessary to change the CCS resistor to draw a little less current?
Zen Mod,
Would motorboating be audible? Is this caused by the power supply capacitance fully discharging during use?
Waiting a bit longer before firing up the SS amp does help. The woofer still moves, but not enough to be concerned about. The "minute" delay I mentioned earlier was just my perception. I would wait until the tubes were glowing then just a little longer. Obviously this wasn't enough time.
Brian,
I would like to build a gas tube regulator as you suggest, but I may have one issue with this. The tranny (Hammond 261G6) I have is rated for 130ma so would be really close to its max current rating. Would it be preferable or necessary to change the CCS resistor to draw a little less current?
Zen Mod,
Would motorboating be audible? Is this caused by the power supply capacitance fully discharging during use?
BGcap said:
BGcap,
Good, I'm glad that a little more waiting for warm-up is all that you needed.
Someone else added the gas tube shunt power supply to my original 12B4A schematic. But, I think it’s a very good power supply choice. The OD3 is specified to work over a current range of 5mA to 40mA. Here it is running at 30mA. You could safely drop that down to 20mA, I would think. As I’d mentioned in prior postings on the 12B4A preamp, I have also tried running the plate current a bit lower than the original 29/30mA with success. It also seems to work quite well at 22mA and 25mA. I’ve varied plate load resistors from 6.2K, 6.8K, 7.5K and 8.2K (all power wire-wound) depending on plate current and the B+ voltage chosen.
If you use separate gas regulators for each channel, and dropped the OD3 current to 20mA, and dropped the 12B4A current from 30mA to, say for example, 24 mA, the total DC current drawn from the raw supply would decrease from 60+60 = 120mA, to 44+44 = 88mA. Of course, you must increase the value of the resistor between the LM317 ADJ and OUT pins using the formula Rset = 1.25/Iset (approximately).
BTW, since this is a class A design, we could replace (or augment) the dropping resistor that feeds the OD3s and the plate resistor with a series current source. Lynn Olson does this in his amps. Power supply noise rejection is then enormous. Our simple little line stage is now getting more complicated, but I think the additional effort and cost might be worthwhile.
I have the "very simple" version up and running, and am very impressed. I like it so much in fact, that I am having trouble using my old preamp for anything but vinyl. Has anyone here ever implemented a phono equal. stage on this guy? I think it would mate well with the V-15s I use which have a slightly, lower than most output. The 12B4 seems to deliver tons of gain. Jay
BGcap said:
I agree in everything with B. Beck ;
you can decrease current through OD stack to some 20 mA,but I must tell that I choose 30 mA during construction (if that little schmtc of shunt reg is worth calling "construction" ) counting also on worst case of -10% of mains variation.even if in my place I have no this troubles ages ago,I still have same habit when I try to calculate PSUs........so-if you use less than 30mA in this case,if mains go near -10% gas toob will probably just fade with more or less loud "plop" from your spks......not nice event...
anyway-try to Google for motorboating and everything will be clearer to you; in this moment I can't find any other english word than-motorboating is criticall sagging of voltage,when RC constants are not adequate for current demands of stage or stages.
in case when you have more than one stage -motorboating is same,but then clearly explained as "not enough current decoupling among stages".........
I hope that is helpfull.
I'm still for increasing at least last cap in chain ......
Gas regulator basics
For those of us trying to understand the voltage regulator part if this project, I found this site to be of some help.
http://www.electronixandmore.com/articles/pwrsupplies.html
Scroll down to the voltage reg. section. It's the basics, but does help.
Ryan
For those of us trying to understand the voltage regulator part if this project, I found this site to be of some help.
http://www.electronixandmore.com/articles/pwrsupplies.html
Scroll down to the voltage reg. section. It's the basics, but does help.
Ryan
Motorboating is usually associated with multi-stage amplifiers. An undesirably high shared power supply impedance, such as when the final PS cap no longer provides a low impedance to ground at very low frequencies, creates unintended feedback paths between stages. Here we've got but one stage per channel, so there is no feedback between stages. One channel can "talk" to the other at very low frequencies if they share a supply but they won't sustain an oscillation. Since this is a richly-biased class A design, power supply signal currents are small compared to the DC current flow anyway.
Thanks again for the help, Brian.
The gas tube regulator looks like the way to go. Less expensive than the better SS supplies and much easier to build as a point to point circuit. And of course the cool factor of the glowing tubes.
There are some who suggest using a capacitor across the gas tubes and some who don't. There is also the idea of putting a resistor across one of the tubes to make sure one lights before the other. I have read about this, but what have some of you who may have built this found works best? Also how important is the quality of the capacitor in this power supply?
The series current source is an interesting idea. I went to Gary Pimm's website to have a look. This may be a good upgrade path after building the gas tube regulated supply using resistors. Probably not very expensive either. Just a small PCB with a handful of what look to be cheap components.
This gas tube can be bypassed with a capacitor value of up to 0.1uF. Beyond that value the OD3 can become a relaxation oscillator. With two gas tubes in series, you could bypass both tubes with one 0.05uF cap (or 0.047uF) maximum, which is the series equivalent of two 0.1uF caps, one across each tube. Yes, this cap will dominate the power supply characteristics at the top of the audio band. So, yes, it should be a nice quality film cap. I have seen some folks add a 150K resistor across the lower gas tube in a two tube series stack, but the data sheets that I have do not mention this point, and a vintage laboratory-grade regulated power supply that I use has a pair of OD3s in series without this “dis-equalizing” resistor. But it can’t hurt.
The series current source option will greatly improve power supply noise rejection. With a simple series resistor, the resistance value is not so much greater than the dynamic resistance of the two gas tubes in series. The resultant voltage divider ratio, which is applied to the incoming hum levels, will not be very great, so you would have to have a very quiet supply ahead of the resistor.
The series current source option will greatly improve power supply noise rejection. With a simple series resistor, the resistance value is not so much greater than the dynamic resistance of the two gas tubes in series. The resultant voltage divider ratio, which is applied to the incoming hum levels, will not be very great, so you would have to have a very quiet supply ahead of the resistor.
Since I have some 12B4's and most of the associated parts, I'll most likely be building one of these soon. The gas regulator tubes look cool enough that I'm going to have to try it with those.
For anyone interested, I found a seller on ebay who seems to be selling NOS JAN OD3's for $1.79 each:
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=120051220807
I'm not associated with the seller, but after poking around for these tubes, this seems like a pretty good deal.
For anyone interested, I found a seller on ebay who seems to be selling NOS JAN OD3's for $1.79 each:
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=120051220807
I'm not associated with the seller, but after poking around for these tubes, this seems like a pretty good deal.
I am interested in monkeying with the 12B4 for audio pre...
This thread got my interest sparked... I just got back from the electronics store...They had a ton of these tubes... I bought a RCA and I also bought a Sylvania JAN spec one in a white box...the writting looks to be green from the 50's or early 60's... Does anyone know which one to get or avoid...I hope these are not microphonic... They were $12 each...
I will report my findings....
Chris
This thread got my interest sparked... I just got back from the electronics store...They had a ton of these tubes... I bought a RCA and I also bought a Sylvania JAN spec one in a white box...the writting looks to be green from the 50's or early 60's... Does anyone know which one to get or avoid...I hope these are not microphonic... They were $12 each...
I will report my findings....
Chris
Ok, having decided to start gathering parts to build a shunt gas tube reg, I would like to ask for opinions:
How important is the quality of the resistor used to set the voltage?
What is a good value for the load resistor?
Is a J/J 100uF electrolytic a good choice for the filter cap?
Preferences for SS diodes? 1n4007, FR107, etc?
Thanks
How important is the quality of the resistor used to set the voltage?
What is a good value for the load resistor?
Is a J/J 100uF electrolytic a good choice for the filter cap?
Preferences for SS diodes? 1n4007, FR107, etc?
Thanks
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.