Reviewing the CL thermistor specs I am a little lost as to what will work for a mono PP EL84 amp.
Just looking for a slow turn on. Any idea where to start? Are they generally reliable?
Just looking for a slow turn on. Any idea where to start? Are they generally reliable?
No, I want a slow turn on of the B+ voltage. The thermistor will be placed in series after the B+ bridge rectification. I should have been more clear.
Go to Ametherm, or google surge supressor calculator. You will need to determine what your greatest surge problem will be, likely power supply capacitance. You will also need steady state current/wattage of amp, peak line voltage. In general not knowing anything about your amp you are aiming for a 5 to 10 ohm, 240 volt rated, 3-5 amp ntc thermistor. It will slow down surge until it heats up but we are talking seconds not minutes. My mc275 with a cl-40 takes 5-6 seconds to produce music. The problem is it stays hot, if the power flicks on and off, or someone switches it rapidly on and off it provides no protection. I add a timed relay to my homebrews that delays AC to B+ and when closed takes thermistor out of circuit. But then you have to figure out away to have heaters and bias [n first. Back to McIntosh tube designs. No fancy relays just that NTC and they are rock solid. I believe you are looking for a cl-150/160 but do the math.
What you want is what I repaired for many decades and was connected to the series tube heater chain in ( dangerous ) radios without mains transformers.
I am looking at an old GEC circuit with a Barretter type 504 in series with the heater chain , this was replaced in later years with a old school Thermistor .
There has been so much talk on DIY Audio about this that I dug out a packet of original RS Thermistors -TH-1A used in those types of radios , big black "resistor looking " shape.
They do work but injected noise --never fear this is 2021 and according to this US prestige engineering website all is now --"well " ?
Looking through my full library collection of Radio & Television Servicing I cannot find any directly feeding the radio tube plates (anodes ) .
Noise characteristics of thermistors: Measurement methods and results of selected devices: Review of Scientific Instruments: Vol 88, No 2
I am looking at an old GEC circuit with a Barretter type 504 in series with the heater chain , this was replaced in later years with a old school Thermistor .
There has been so much talk on DIY Audio about this that I dug out a packet of original RS Thermistors -TH-1A used in those types of radios , big black "resistor looking " shape.
They do work but injected noise --never fear this is 2021 and according to this US prestige engineering website all is now --"well " ?
Looking through my full library collection of Radio & Television Servicing I cannot find any directly feeding the radio tube plates (anodes ) .
Noise characteristics of thermistors: Measurement methods and results of selected devices: Review of Scientific Instruments: Vol 88, No 2
Last edited:
Look at the 555 based delay in my free gerbers thread, This is an electronic solution to your problem. I can also provide kits, but shipping can be a PITA.
With the gerbers you can order the boards in china yourself.
Provided the AC from the HV transformer isnt higher than ~350VAC you can use a finder 41.51 series relay to switch the HV on at the AC side after about 30 sec.
Cheers,
V4lve
With the gerbers you can order the boards in china yourself.
Provided the AC from the HV transformer isnt higher than ~350VAC you can use a finder 41.51 series relay to switch the HV on at the AC side after about 30 sec.
Cheers,
V4lve
I use a NTC in one of my regular use amplifiers.
It is a push-pull EL84 too, but two channels instead of mono.
So far, so good. Of course, I try to never turn power off and then on again in a short time.
The NTC I'm using is a 60 ohm cold resistance, and aprox 15 mm diameter.
It probaly is a few ohms when hot. Even if it is 10 ohm when the amplifier is operating normally, the wasted power in heat is not a big deal.
Maybe you could try EPCOS B57237S0600M0, which seems very close to what I'm using.
Or, maybe better, as your amp is got a single channel, B57237S0330M0, which is half the resistance.
ontariomaximus,
Soft Start B+:
If you have a 6.3V filament secondary, a B+ secondary that has plenty of voltage, and a 3rd (a separate 6.3V filament winding), then try the following intrinsic B+ soft start:
Using the separate 6.3V secondary, a series 0.65 Ohm resistor, and a 5Y3.
The 5Y3 has a 5 Volt, 2 Amp filament is 2.5 Ohms when the filament is warmed up.
But the filament is far less than 1 Ohm (probably close to 0.65 Ohm) when the filament is cold. The filament will warm up very slowly, and that will delay the B+ start-up.
A type 80 rectifier will work just as good as a 5Y3 there, and give you a shoulder tube to look at.
Or, if you have a 5AR4, and a 5V filament winding, the indirect heated cathode takes time to warm up.
Soft Start B+:
If you have a 6.3V filament secondary, a B+ secondary that has plenty of voltage, and a 3rd (a separate 6.3V filament winding), then try the following intrinsic B+ soft start:
Using the separate 6.3V secondary, a series 0.65 Ohm resistor, and a 5Y3.
The 5Y3 has a 5 Volt, 2 Amp filament is 2.5 Ohms when the filament is warmed up.
But the filament is far less than 1 Ohm (probably close to 0.65 Ohm) when the filament is cold. The filament will warm up very slowly, and that will delay the B+ start-up.
A type 80 rectifier will work just as good as a 5Y3 there, and give you a shoulder tube to look at.
Or, if you have a 5AR4, and a 5V filament winding, the indirect heated cathode takes time to warm up.
Last edited:
Elerion,
Turning the power off, and then on again quickly can not be prevented . . .
. . . if there is a brief power mains interruption, and quick mains on again event (called a hot start).
Either design the amplifier to take hot starts gracefully; or use your own personal home 'faultless' mains power supply generator.
Turning the power off, and then on again quickly can not be prevented . . .
. . . if there is a brief power mains interruption, and quick mains on again event (called a hot start).
Either design the amplifier to take hot starts gracefully; or use your own personal home 'faultless' mains power supply generator.
ontariomaximus, as per Post #4, you need to identify all the operating conditions of your amp that will be influenced by the NTC. If you can't do that then I suggest you don't add an NTC that could be inadequately designed.
Btw, you haven't elaborated on why you want B+ rise time to be slowed, or if that also includes a delay.
Btw, you haven't elaborated on why you want B+ rise time to be slowed, or if that also includes a delay.
good idea - I could even try a TV damper diode like a 6AX4, but I think I looking for a simple and small solution
Just a slow ramp up to protect the caps and maybe lessen the shock to the switch. The unloaded PS voltage is quite high.
Does that mean you are just going to guess at a part to use?
How are you going to measure the change and if it is effective both during start-up, and for continuous operation ?
How are you going to measure the change and if it is effective both during start-up, and for continuous operation ?
Use capacitors with voltage ratings that are higher than the unloaded voltage of the B+.
Use a switch that is rated for more than the maximum turn-on Inrush current.
Simple, always works.
Or,
Use parts that are poorly rated, and expect poor results.
As my high school Physics teacher used to say, "Always, Forever, Eternally, Amen".
Use a switch that is rated for more than the maximum turn-on Inrush current.
Simple, always works.
Or,
Use parts that are poorly rated, and expect poor results.
As my high school Physics teacher used to say, "Always, Forever, Eternally, Amen".
Patrick Turner devised some kind of soft start circuit several years ago. It included a complete reset after turn off to avoid hot switching if immediately switched on again. Probably find it still on his website.🙂
But a CL delay should be OK on the B+, their largest application is use on the front end of switch mode power supplies.
I've used several on the line side of some of my amps, worked well.
But a CL delay should be OK on the B+, their largest application is use on the front end of switch mode power supplies.
I've used several on the line side of some of my amps, worked well.
True. There're better approaches. But IMHO, if someone is after a simple (not bulletproof) solution, a NTC is something to be considered.
Analog based time delayed realy switching is probably a good choice.
Some people use digital electronics for these tasks, but some solutions might not be bulletproof either... especially those who use a small microcontroller which can hang for so many reasons 😀
Elerion,
Nothing is completely bullet proof.
Fuses have been known to "open" and sustain a plasma arc long enough for the rest of the circuit to be damaged.
If I remember correctly, Apollo 13 started with 24V heaters, and were changed to 65V heaters. But the overheat safety switch was not rated for 65V.
But perhaps that is just an old wives tale.
What I hate to see in an amplifier is 3 added parts to different circuits in the amp, or 3 different protection circuits that only addresses 3 failure modes,
Versus . . .
using properly rated parts that addresses 5 different failure modes.
I think thatt using parts that are rated for the worst case abuse they would receive, is the simplest, most fail-safe, and generally the best solution.
"You should make things as simple as possible, but no simpler." Albert Einstein.
Nothing is completely bullet proof.
Fuses have been known to "open" and sustain a plasma arc long enough for the rest of the circuit to be damaged.
If I remember correctly, Apollo 13 started with 24V heaters, and were changed to 65V heaters. But the overheat safety switch was not rated for 65V.
But perhaps that is just an old wives tale.
What I hate to see in an amplifier is 3 added parts to different circuits in the amp, or 3 different protection circuits that only addresses 3 failure modes,
Versus . . .
using properly rated parts that addresses 5 different failure modes.
I think thatt using parts that are rated for the worst case abuse they would receive, is the simplest, most fail-safe, and generally the best solution.
"You should make things as simple as possible, but no simpler." Albert Einstein.
Last edited:
