UK Vintage Radio Repair and Restoration Discussion Forum

UK Vintage Radio Repair and Restoration Discussion Forum (https://www.vintage-radio.net/forum/index.php)
-   Components and Circuits (https://www.vintage-radio.net/forum/forumdisplay.php?f=40)
-   -   Fan speed control by OP of amp. (https://www.vintage-radio.net/forum/showthread.php?t=159124)

Diabolical Artificer 22nd Aug 2019 10:28 am

Fan speed control by OP of amp.
 
1 Attachment(s)
The two big monoblocks I built get very hot after a while, I have a fan drawing air up through the PSU, this then goes into the amp chassis to cool things there, then through six holes round each OP valve. To reduce fan noise I put a 100r to slow the fan down, at low volume it can't be heard, but to cool the amp it needs to be on full whack.

So I thought of taking the OP after the OPT, into a buffer, a 10M and Opamp buffer, into a diode to rectify, into a "comparator", OP from this to a tranny to turn on a relay which will short out the dropping R on the fan, see att.

Anything I've missed? Andy.

short wave 22nd Aug 2019 10:40 am

Re: Fan speed control by OP of amp.
 
Diodes will clamp O/P across speaker (0.6 volts if silicon ones used) ...distortion at high volume

vidjoman 22nd Aug 2019 10:51 am

Re: Fan speed control by OP of amp.
 
Why not just a simple thermostat?

dave cox 22nd Aug 2019 11:24 am

Re: Fan speed control by OP of amp.
 
I think you missed the series resistor (as short wave points out)!

Better to monitor the temperature though, it could get annoying with the fan clicking on and off if you listen to something with a high dynamic range. Maybe use a low voltage DC fan and use proportional control ?

dc

vidjoman 22nd Aug 2019 11:50 am

Re: Fan speed control by OP of amp.
 
You could use 3 x 33 ohms in series with a thermal switch set to different temperatures across each resistor. That’ll give you 4 speeds getting progressively faster if the temperature rises too much.

mhennessy 22nd Aug 2019 12:38 pm

Re: Fan speed control by OP of amp.
 
1 Attachment(s)
The diode clamp (with a series resistor) stops the output of the first op-amp going above 0.6V.

This means the inverting input to the second op-amp will never go about 0V.

As the non-inverting input is at some higher voltage, the output of the op-amp will be high, and the relay will be on. So the fan runs at high speed all the time...

That ignores the missing resistor between op-amp and transistor, which I'm sure you planned to add. If the op-amp's negative connection is 0V, then you'll need another resistor so that the transistor can turn off. If it's -ve by more than about 7V, then you'll need a protection diode at the transistor base. Also, it's not good practice to have nothing to earth after that series diode - if the op-amp uses PNP input transistors, then there's nothing to provide input current. A resistor to ground is all that's needed.

Another important thing - suppose you resolve all these issues (which is easy enough), there's no time constant. So the relay coil will buzz on and off in time with the music. It will act like a miniature loudspeaker. So a cap after the rectifying diode is needed at the very least.

If it's a DC fan, then you don't need a relay - you can drive it in a number of ways, including using an emitter follower or an LM317.

I would also recommend using a temperature sensor rather than a peak voltage sensor.

Over the years I've built many fan speed controllers for PCs, as I can't stand fan noise. Usually just 2 transistors and a 10k NTC thermistor.

Attached is a more sophisticated version that has 2 preset controls - one to set the default "idle" speed when the temperature is low, and another one to decide at what temperature the fan starts to speed up.

Driving the fan is a CFP that has a gain of nearly 4, so the voltage at the base varies between approximately 0 to 3 volts. Using a CFP means that you can get almost the full 12V at the fan. Also note the cap in series with the 10k resistor - this "kicks" the fan at power-on, which is good when the fan bearings get old, and also if you want the idle speed to be lower than that needed for the fan to start up.

This is for a case fan mounted on the front of the box. The NTC thermistor is located on an exit grille - as the exhaust air warms up, the intake fan speeds up. It works really well for such a simple circuit.

Apologies for the crudity of the sketch, but I haven't got around to writing it up (it's not really worthy of a website article). If you want to have a play, I'll happily help.

Cheers,

Mark

Diabolical Artificer 22nd Aug 2019 1:16 pm

Re: Fan speed control by OP of amp.
 
Thanks chaps, on reflection my idea was rubbish,the schematic was just a rough back of a fag packet styley with no R/C's etc. Temp control would better as pointed out, I did pull a simple two tranny + NTC off a ATX PSU, but that circuit looks better Mark. However as the amp gets bl**dy hot after about 30 mins, this means the fan will be on full whack pretty much all the time, hence original idea.

I think I'm going to have to live with the fan, which to be honest you can't hear with music on as I mostly play rock, I wonder what it would be like with classical? Will give it a whirl.

Thanks for the ideas/FB etc, Andy.

mhennessy 22nd Aug 2019 2:13 pm

Re: Fan speed control by OP of amp.
 
There's no doubt that a fan running continuously is easier to ignore than one that stops and starts.

The bigger the fan, the more airflow it moves for a given rotational speed - I'm really impressed with the large 120mm case fans, that can run at much slower speeds for a given airflow than the traditional 80mm fans that all computers used back in the old days. If you can make them fit, then they will definitely give you more cooling for less mechanical noise than the smaller ones.

Things like rubber mounts (even just grommets) help to avoid conducting motor noise into the chassis, and keeping air speeds low to avoid turbulence will make a big difference. Anything you can do to improve airflow really helps - it might mean enlarging holes, but for safety it might be wise to put a mesh grille to keep fingers out?

I'm sure it'll be less of an issue in the winter ;)

Refugee 22nd Aug 2019 2:18 pm

Re: Fan speed control by OP of amp.
 
I have found that the fans used in gambling machine cabinets are nice and quiet even at full pelt.
I have a Corsair 140mm fan in my home brew beer cooler and it is very quiet. Its 12V 100ma power supply shares its mains input with a heat pump that is no more noisy than a domestic fridge. The fan is hardly any louder than the pump motor.
If you can find any head room in your cabinet a bigger fan would be the way to go.

Biggles 22nd Aug 2019 10:03 pm

Re: Fan speed control by OP of amp.
 
Have you considered using a peltier effect module? Nice and quiet and probably easy to control proportionally by adjusting current flow through it. May be a bit on the expensive side unless you can find a couple in one of those small 12V beer cooling mini-fridge gadgets. Just an alternative approach to the problem.
Alan.

McMurdo 22nd Aug 2019 10:40 pm

Re: Fan speed control by OP of amp.
 
quite a few fans have built in speed control, Papst do a few, look in rs/farnell etc. You add a thermistor in an appropriate place and attach it to the fan control wire. The thermistor is usually included.
You will have to search the datasheets as its rarely mentioned in the shortform spec on the product page.

Diabolical Artificer 23rd Aug 2019 7:08 am

Re: Fan speed control by OP of amp.
 
Good points there Marc, though adding a bigger fan at this stage will need quite a re-build, I have however drilled holes out to increase airflow, will do some tests today.

Again, Simon, fitting a bigger fan + power supply is going to be megga tricky, but take your point.

That's an idea Alan, just watched the lateset Electroboom vid where he mucks about with the things and also mentions fan turbulance.

Thanks Kevin, didn't know that, something to bare in mind.

Going back to the original idea, there are a few things I thought of to make it less of a pain, 1) to make it latching or 2) put a "bucket" in there somewhere so fan isn't off/on all the time. Temp control is still preferable though, just reporting what the noggin came up with while doing the washing up/hoovering : )

Andy.

McMurdo 23rd Aug 2019 9:41 am

Re: Fan speed control by OP of amp.
 
1 Attachment(s)
Just had a quick look for the record, papst part numbers have a 'V' at the end of the model number for externally connected thermistor (3-wire), or 'I' at the end for an on-board (in the airflow) thermistor (2-wire).

Herald1360 23rd Aug 2019 12:24 pm

Re: Fan speed control by OP of amp.
 
Surely using the amp output to control the fan is a bit pointless. Unless the amp is running pure class B, the power it wastes as heat isn't going to change much whether it's idling or at full chat.

Commercial 100W or so valve amps would have relied on natural convection- is this homebrew a shoehorn job regarding layout and cabinet? How hot do critical heat sensitive bits like electrolytics and wound components actually get? They'll be rated for operating temperatures that are very "hot" by the standard of a finger poke test!

mhennessy 23rd Aug 2019 1:02 pm

Re: Fan speed control by OP of amp.
 
That's why I like infra-red thermometers :thumbsup:

They sometimes pop up in Lidl and Aldi for just over a tenner. I bought one to see how it compared to my Fluke (bought long before you could get them at these prices), and it's really not bad at all.

Haven't gone for an IR camera yet, but if I ever win the lottery ;)

Biggles 23rd Aug 2019 6:01 pm

Re: Fan speed control by OP of amp.
 
I recently got an infra-red thermometer from Screwfix (about thirty quid I think) and am well impressed with it, not having had one before. Great for checking things like heatsinks/component temperatures.
Alan

Diabolical Artificer 23rd Aug 2019 6:28 pm

Re: Fan speed control by OP of amp.
 
Yep, an infra red thermo would be good but having to really on Mk1 fingers. Yes Chris, it is bl**dy pointless, my goal was to quiet the fan at low volume levels, however I've had them on today for a few hours at moderate volume and they get well hot.

First, #1 say was run with the fan full on, top chassis "plate" hot, but not "ouch" hot. Mains tfmr inside hot but not painfull to the touch. Fan noise noticeable, but not at moderate volume 4ft away.

#2 I ran with two 100r in parallel to reduce fan speed, top chassis plate not "ouch" hot, but mains tfmr isn't touchable. Fan noise not noticeable close to even at zero volume.

The fans were an after thought sort of, no thermal design as such, didn't realise amps would get so hot. Mains tfmr's were wound such that toroidal cores were more than adequate, IE 500VA actual power, 625VA cores.

I think I'm going to have to re-fettle chassis and bottom plate so as to fit bigger, quieter fan and better air flow.

Andy.

Herald1360 23rd Aug 2019 7:52 pm

Re: Fan speed control by OP of amp.
 
Sounds like you need some temperature figures for the transformers. Winding temps can be inferred from cold vs hot resistance readings and the tempco of copper and the core can be measured directly. If they're both under 100C the TX should be happy even if your fingers ain't!

Even 60C can produce irreversible tissue damage.....

CambridgeWorks 23rd Aug 2019 9:56 pm

Re: Fan speed control by OP of amp.
 
Andy, I have some new AMD cpu heatpipe heatsinks with fans that have 4 wires.
see: https://www.ebay.co.uk/itm/AMD-HEATS...W/130756917825 which are the same.
Fans are about 80mm square from memory. I can strip a few fans from the heatsinks (to save weight) and post if any use for you? Don't know what 4 wires do, obviously +12, -12, maybe a temp sensor and a loss of rotation sensor??? Pcs are not really my thing!
let me know if any use?
Rob

mhennessy 23rd Aug 2019 10:02 pm

Re: Fan speed control by OP of amp.
 
Usually, the 2 extra wires on 4 wire fans are for tacho out and PWM in.

http://pavouk.org/hw/fan/en_fan4wire.html


All times are GMT +1. The time now is 9:25 am.

Powered by vBulletin®
Copyright ©2000 - 2024, vBulletin Solutions, Inc.
Copyright ©2002 - 2023, Paul Stenning.