Comms PSU

[TomHoward]

I thought that, but I had hoped that the PSU would have some protection. I also have a Sony laptop PSU that I could use instead - although this was marked at 19.5V as opposed to 20. (Whilst they actually both output 19.8 when measured.).

 

With regards to inducing noise, is that if the multi has a common earth? I tested it with the multi but it was the only thing on it at the time, didn't test it with the multi in use. I might suggest that they run an additional cable alongside the multi for comms use.

[TomHoward]

Just to add to this that this just got failed by our external PAT testers, as the PSU says on it in big letters that it "Must only be used with Dell computer equipment", so in retrospect I should have used a generic / off the shelf power supply.

[mutley]

Just to add to this that this just got failed by our external PAT testers, as the PSU says on it in big letters that it "Must only be used with Dell computer equipment", so in retrospect I should have used a generic / off the shelf power supply.

 

IMHO that is not a valid reason to fail it. It the label was missing would they fail it? A "PAT" test is an Electrical Safety Test. Is the electrical appliance safe to use?

The only reason that is stated on the label, is the same reason the third wire is - Dell's awkward greed in forcing you to use Dell parts for everything. (The third wire - the little one - is to tell the laptop what model charger it is, its serial number and its power rating. It does this in serial data along the thin wire normally connected to the centre pin. Dell batteries employ a similar tactic)

There is NOTHING safety-wise to stop you using that power pack for any other purpose. That's like saying "Only use genuine Sennheiser power supply unit" on a radio mic receiver. Is it dangerous to use a different supply (or the same rating and polarity)? Of course not!

Take the FAILED label off and use it. As has been previously explained, PAT testing itself is not a legal requirement anyway! (only to maintain all electrical equipment regularly to a safe standard and make accurate records of any maintenance carried out)

[TomHoward]

I have just pulled the 'only Dell' sticker off it and carried on. Just thought it was worth mentioning in case anyone else was going to copy it - I can see the valid complaint from the tester though, if he's just a box-checker one of the questions would be 'is it being used as the manufacturer intended?' and this would have to be a no. Passing it might require some thought as to whether the new use was safe..

[sunray]

Well I managed to do a quick test on a Techpro PS711. It's rated at 24V, 0.64A (640mA).

 

Open circuit voltage =25.2V

 

2 12V 6W bulbs in series is theoretically 0.5A (500mA)

 

measured as 430mA & 20.8V

 

My calculation is a 4.4V drop at 430mA = 10.23Ohms

 

I assume (always a silly thing to do) that the 24V supply is of negligible output resistance and R1 in the above circuits is 10 Ohms, equally it could be 3.3 Ohms and the psu drops 3V doing it's current limiting.

 

The PS711 is claimed to operate 16 headsets so they should operate at 40mA which doesn't quite match with my earlier recollections of 6 drawing 400mA (66mA) so I assume I'd measured the whole system and not just the 6 on the end of a long run.

 

I resisted the temptation to cut the warranty void sticker to have a look.

 

330 Ohms at 640mA would be a voltage drop of 211V which can not happen from a 24V supply.

 

 

 

[richardash1981]

Well I managed to do a quick test on a Techpro PS711. It's rated at 24V, 0.64A (640mA).

 

Open circuit voltage =25.2V

 

2 12V 6W bulbs in series is theoretically 0.5A (500mA)

 

measured as 430mA & 20.8V

 

My calculation is a 4.4V drop at 430mA = 10.23Ohms

 

I assume (always a silly thing to do) that the 24V supply is of negligible output resistance and R1 in the above circuits is 10 Ohms, equally it could be 3.3 Ohms and the psu drops 3V doing it's current limiting.

 

The PS711 is claimed to operate 16 headsets so they should operate at 40mA which doesn't quite match with my earlier recollections of 6 drawing 400mA (66mA) so I assume I'd measured the whole system and not just the 6 on the end of a long run.

 

I resisted the temptation to cut the warranty void sticker to have a look.

 

330 Ohms at 640mA would be a voltage drop of 211V which can not happen from a 24V supply.

The 330R series resistor does look very large in that circuit. The beltpack power consumption will vary wildly depending on whether the line is silent or someone is ranting (because the headphone drivers are class B), so I wouldn't bank on measurements being very consistent. This is another reason for thinking that a large value series resistor is a bad idea - the supply voltage will vary a lot as the audio signal does.

 

For a linear, unregulated, PSU (I assume what the TechPro PS711 is) then the 10 ohms output impedance is not unreasonable. However I can't see any reason why this is necessary to the operation of the system - unlike the audio network, the intercom should work just fine with a stiff (low source impedance) supply rail. With the 330R resistor shown the short-circuit current will only be 73mA which is not likely to operate more than one or two beltpacks!

 

The only reason I can see for R1 at all is that it forms a low pass filter with C2 to remove noise from the power supply, which will tend to get into the audio line. However it need to be a much smaller value part in order to keep the DC conditions sensible. I think this is a decimal place error - a 3.3 ohm, resistor will keep the voltage drop to ~2V which is OK, and with a load current of 0.64A flowing, the power dissipation in the resistor is 1.35W, which is just about OK for a 2W part. In practice the dissipation will be lower because of the class B beltpacks and peaky nature of audio signals, so the resistor should be fine.

 

Conclusion: anyone trying to build a power supply to the schematic by John Baraclough needs to make R1 = 3.3 Ohms (two orders of magnitude smaller than shown), or (if they have a low noise external PSU, e.g. a linear regulator) omit it altogether.

[sunray]

There needs to be something in there for current limiting and cable safety, if a straight forward 24V 2A supply were to be used without it there is a risk of a short circuit melting a cable as there is a distinct possibility of 10A undre such conditions.

 

Well I managed to do a quick test on a Techpro PS711. It's rated at 24V, 0.64A (640mA).

 

Open circuit voltage =25.2V

 

2 12V 6W bulbs in series is theoretically 0.5A (500mA)

 

measured as 430mA & 20.8V

 

My calculation is a 4.4V drop at 430mA = 10.23Ohms

 

I assume (always a silly thing to do) that the 24V supply is of negligible output resistance and R1 in the above circuits is 10 Ohms, equally it could be 3.3 Ohms and the psu drops 3V doing it's current limiting.

 

The PS711 is claimed to operate 16 headsets so they should operate at 40mA which doesn't quite match with my earlier recollections of 6 drawing 400mA (66mA) so I assume I'd measured the whole system and not just the 6 on the end of a long run.

 

I resisted the temptation to cut the warranty void sticker to have a look.

 

330 Ohms at 640mA would be a voltage drop of 211V which can not happen from a 24V supply.

The 330R series resistor does look very large in that circuit. The beltpack power consumption will vary wildly depending on whether the line is silent or someone is ranting (because the headphone drivers are class B), so I wouldn't bank on measurements being very consistent. This is another reason for thinking that a large value series resistor is a bad idea - the supply voltage will vary a lot as the audio signal does.

 

For a linear, unregulated, PSU (I assume what the TechPro PS711 is) then the 10 ohms output impedance is not unreasonable. However I can't see any reason why this is necessary to the operation of the system - unlike the audio network, the intercom should work just fine with a stiff (low source impedance) supply rail. With the 330R resistor shown the short-circuit current will only be 73mA which is not likely to operate more than one or two beltpacks!

 

The only reason I can see for R1 at all is that it forms a low pass filter with C2 to remove noise from the power supply, which will tend to get into the audio line. However it need to be a much smaller value part in order to keep the DC conditions sensible. I think this is a decimal place error - a 3.3 ohm, resistor will keep the voltage drop to ~2V which is OK, and with a load current of 0.64A flowing, the power dissipation in the resistor is 1.35W, which is just about OK for a 2W part. In practice the dissipation will be lower because of the class B beltpacks and peaky nature of audio signals, so the resistor should be fine.

 

Conclusion: anyone trying to build a power supply to the schematic by John Baraclough needs to make R1 = 3.3 Ohms (two orders of magnitude smaller than shown), or (if they have a low noise external PSU, e.g. a linear regulator) omit it altogether.

 

 

 

 

 

[timsabre]

All psu modules are current limited these days and if they say 2A then you won't get any more out of them. In the old days when this circuit was designed with a transformer psu, the current limit resistor would be a good idea. But unnecessary now I think.

[james.sealey]

Apologies for reviving this old topic, but I've found it very useful and am in the process of building my own PSU. I do have a question regarding an 'aux input', ie bringing an external audio signal into the comms ring, is it just a case of sticking it down the audio pin with some form of attenuation for level control?

 

Thanks In Advance

 

James

[timsabre]

Apologies for reviving this old topic, but I've found it very useful and am in the process of building my own PSU. I do have a question regarding an 'aux input', ie bringing an external audio signal into the comms ring, is it just a case of sticking it down the audio pin with some form of attenuation for level control?

 

No, this will knacker up (technical term) the impedance balancing of the audio line. You need to use the audio hybrid circuit from a beltpack, you could use one of the bluecom boards for this.