folks, I'm about to build up a single stage preamp using a triode connected EL84 with a 70H plate choke.
At this point, i'm seriously considering placing a stack of NiMH cells for computer memory backup under the cathode to provide the required bias (approx 10v). The cells are designed for continuous overcharge and a constant trickle of 30ma should do them no harm.
Am I missing some major caveat here. Folks only ever seem to talk about battery bias on input and driver stages, never on things handling this level of current.
Thanks in advance
Drew
At this point, i'm seriously considering placing a stack of NiMH cells for computer memory backup under the cathode to provide the required bias (approx 10v). The cells are designed for continuous overcharge and a constant trickle of 30ma should do them no harm.
Am I missing some major caveat here. Folks only ever seem to talk about battery bias on input and driver stages, never on things handling this level of current.
Thanks in advance
Drew
I strongly recommend putting cells not in cathode network but in grid one (cathode should be grounded in this case) - minus to grid, plus to input and grid leak resistor.
Of course low noise cells are desirable, but even alkaline AA (or AAA) batteries or watch cells (CR21) will work great.
Good luck!
Of course low noise cells are desirable, but even alkaline AA (or AAA) batteries or watch cells (CR21) will work great.
Good luck!
> a stack of NiMH cells for computer memory backup under the cathode to provide the required bias (approx 10v)
You mean like a Zener?
The early NiCads were not good batteries but did work and were used as voltage stabilizers.
BTW: common Alkaline batteries can also be used as voltage stabilizers, and may be a lot cheaper than good NiCad or NiMH. And you could probably use the "dead" batteries from other toys, if you catch them before they go flat-dead. (However this class of battery does tend to have some noise.)
With modern NiCads, when they get old, they can go high-resistance. I have seen over 7 volts while attempting to charge an old "3.6V" NiCad.
If series NiCads are left un-charged with a load, the weakest cell eventually gets reverse-voltage from the other cells in the series-stack and will grow internal shorts. (This is perhaps the most common failure mode in PC type NiCad applications.)
For your case: I suppose high-resistance will do you no great harm, just reduce tube idle current. When it becomes a problem (many years from now) you'll put in new cells.
A shorted cell is more dangerous to the amp; idle current rises. However if the only path out of the cell is through the tube Cathode, then when the tube is ON the cell is charging, and when it is off there is no leakage path (except each cell's own internal leakage). So that may not be the problem it is in most NiCad devices. And worst comes to worst: you burn-up a 6BQ5, not a big-huge deal.
And NiMH is different. I have not worked with NiMH enough to know what is different.
I don't see any reason it won't work. Go ahead and try it, let us know how it sounds and if any problems appear.
Do put the cell some place where if the unexpected happens, and the battery bursts, it won't splash at your face or on the good furniture.
Personally, I think a resistor or Zener and a big fat cap is a better plan, but I'm kinda dull.
You mean like a Zener?
The early NiCads were not good batteries but did work and were used as voltage stabilizers.
BTW: common Alkaline batteries can also be used as voltage stabilizers, and may be a lot cheaper than good NiCad or NiMH. And you could probably use the "dead" batteries from other toys, if you catch them before they go flat-dead. (However this class of battery does tend to have some noise.)
With modern NiCads, when they get old, they can go high-resistance. I have seen over 7 volts while attempting to charge an old "3.6V" NiCad.
If series NiCads are left un-charged with a load, the weakest cell eventually gets reverse-voltage from the other cells in the series-stack and will grow internal shorts. (This is perhaps the most common failure mode in PC type NiCad applications.)
For your case: I suppose high-resistance will do you no great harm, just reduce tube idle current. When it becomes a problem (many years from now) you'll put in new cells.
A shorted cell is more dangerous to the amp; idle current rises. However if the only path out of the cell is through the tube Cathode, then when the tube is ON the cell is charging, and when it is off there is no leakage path (except each cell's own internal leakage). So that may not be the problem it is in most NiCad devices. And worst comes to worst: you burn-up a 6BQ5, not a big-huge deal.
And NiMH is different. I have not worked with NiMH enough to know what is different.
I don't see any reason it won't work. Go ahead and try it, let us know how it sounds and if any problems appear.
Do put the cell some place where if the unexpected happens, and the battery bursts, it won't splash at your face or on the good furniture.
Personally, I think a resistor or Zener and a big fat cap is a better plan, but I'm kinda dull.
Many thanks for the input folks.
There is a 100k pot effectively as the grid leak resistor. Apart from a small 470r resistor on the grid pin itself, the impedance to ground from the grid varies according to the volume setting. Quiet listening with the volume turned down low (and therefore a small resistance to ground from the attenuator wiper) would drain the cell fairly quickly. Certainly faster than 100k would.
I'll have a trial with NiMH on the cathode, on the grid with cathode grounded and using the standard cap/resistor cathode bias and report back.
Drew
There is a 100k pot effectively as the grid leak resistor. Apart from a small 470r resistor on the grid pin itself, the impedance to ground from the grid varies according to the volume setting. Quiet listening with the volume turned down low (and therefore a small resistance to ground from the attenuator wiper) would drain the cell fairly quickly. Certainly faster than 100k would.
I'll have a trial with NiMH on the cathode, on the grid with cathode grounded and using the standard cap/resistor cathode bias and report back.
Drew
Trickle, trickle little star, how I wonder Watt you are?
Bear in mind that you need to choose cells that will tolerate your 30mA as a trickle/continuous charge current, so that means you need cells rated at 300mA/hr. Given that the limit is set by internal temperature effects, and you're going to put them in a hot place, 500mA/hr might be better.
Bear in mind that you need to choose cells that will tolerate your 30mA as a trickle/continuous charge current, so that means you need cells rated at 300mA/hr. Given that the limit is set by internal temperature effects, and you're going to put them in a hot place, 500mA/hr might be better.
Drew
You seem to have got this wrong. The grid circuit consists of a battery and a series resistor (100k) between grid and ground. The pot won't change anything as far as draining the battery is concerned but will unfortunately vary the grid bias. A simple solution will be to incorporate the battery in the ground connection of the pot and forget about the extra resistor. Make sure you listen to this as it sounds different and IMO better than cathode bias.
Sorry, my fubar.
Let me get this straight:
battery goes between ground and 100k resistor, top of 100k resistor goes to input line on one side(with volume pot etc), and is connected to grid via oscillation prevention stopper (470r) on the other side.
Apologies, I had visualised the cct with the 100k being the input load resistor and the cells connected directly between input signal line and ground.
Would be easier with diagram but I think I've got it straight now.
Drew
Let me get this straight:
battery goes between ground and 100k resistor, top of 100k resistor goes to input line on one side(with volume pot etc), and is connected to grid via oscillation prevention stopper (470r) on the other side.
Apologies, I had visualised the cct with the 100k being the input load resistor and the cells connected directly between input signal line and ground.
Would be easier with diagram but I think I've got it straight now.
Drew
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.