COP 17 Heater | Rosemary Ainslie

[poynt99]

Be Reasonable

Yesterday I sent Aaron a PM with that configuration and CC'd you

I guess I'll have to take the time to do the Spice model myself as .99 keeps dancing around my request. It may take a couple of days with everything on my plate - but I'll let you know. Now if I could just get everyone to comprehend my prior posts I think we may be getting to the real answers everyone can agree on.

Cheers,

In all fairness, I have been trying to translate your requests into "SPICE-talk", not dancing around your request. I've explained fairly clearly I think what you've asked so far (voltages), and what further info needed to complete the request (currents and power dissipations).

I'm certain that Protel SPICE is not too different than PSpice, therefore you will have a similar challenge there.

If you would kindly capitulate my requests for the information I need, I can save you a lot of time. I've put a fair bit into your sim and others already on your and everyone's behalf, so am I being unreasonable in asking for these specifics?

.99

[Harvey]

Hi Harvey - Aaron's been doing some tedious control tests that he wanted to get under the belt.

MileHigh has posted a suggestion Post 1608 across the divide. Could you please check this out and see if it works. And if so - then explain it for me or Aaron and he can do those numbers you need.

Incidentally there's a definite 'truce' happening all over the place. A good thing I might add.

Yesterday I sent Aaron a PM with that configuration and CC'd you

I guess I'll have to take the time to do the Spice model myself as .99 keeps dancing around my request. It may take a couple of days with everything on my plate - but I'll let you know. BTW, it's kinda nice to see Mile High catching up to a conversation I had with TK weeks ago regarding the water hammer - Now if I could just get everyone to comprehend my prior posts I think we may be getting to the real answers everyone can agree on.

Cheers,

[witsend]

bridging the divide.

Hi Harvey - Aaron's been doing some tedious control tests that he wanted to get under the belt.

MileHigh has posted a suggestion Post 1608 across the divide. Could you please check this out and see if it works. And if so - then explain it for me or Aaron and he can do those numbers you need.

Incidentally there's a definite 'truce' happening all over the place. A good thing I might add.

[Harvey]

...
I have a concept for driving a resistor (inductor) I thought you might want to look at. Might be of use, maybe not. Or if you want to improve on it (eg to sense saturation using a differential op-amp mode). I hand drew it.
Imageshack - opampcmosinverterconcep

Andrew

Hi Ash, Andrew:
See Figure 11 Here for an example of an Op-Amp drive.

Cheers,

Harvey

[Harvey]

Quote from page 30 of the scope manual:

Ensure that all probe common leads are connected to the same voltage.

.99

Well there 'ya go then - the requested test won't be done.

[Harvey]

I know Aaron isn't reading my posts, and forgive me for stating the obvious, but shouldn't it be determined for certain if the probe grounds are separable or not? I'm not saying it isn't possible, but I've not seen a scope yet that has this capability. Not having the grounds separated electrically and attaching them to different places in the circuit could lead to some very undesirable effects.

.99

You may recall that TK stated the 'Fluke-O-Scope' did in fact have isolated inputs.

[poynt99]

Hi ALL.

Thanks so much Poynt/guys, just got in. Okay here is what i got back from Andrew. We can change the circuit and reproduce Harvey's/ Aaron's new if needed what ever it takes . Poynt got the cop circuit in the pdf and will get Andrew to get onto it my friend.

Ash, the MRA has a rectified/DC output built in to the circuit, whereas the Ainslie circuit does not, so the COP circuit in the PDF won't work right out of the box. A number of posts back I made two suggestions to possibly get around this problem so the COP circuit could be used. One was to use a transformer on the load resistor which feeds a diode bridge, and the second was to wrap a secondary coil on the load resistor itself and use that to feed the bridge.

---------------------

Hey Poynt99,

I've measured the inductance of the resistor - comes out to about 20uH. The meter I have is not very accurate as its main function is a multi meter. I'll look at getting a proper LCR meter though.

That value sounds about right Andrew. It seems most everyone's are 2 or 3 times the inductance than Rosie's was.

The apparent duty cycle of 30% was the smallest I could get with the resistor / capacitor combination specified for that particular 555 timer circuit. I have observed that when you change the duty cycle on a 555 timer, it also changes the frequency. The scope I believe only has a bandwidth of about 10MHz.

I'm surprised that you even got the duty down that low. The lowest I could get was about 70%.

I also observed that with the gate resistance set to 500 ohms, that the amp draw went up. The amp draw was minimum with no gate resistance (see photos).

Yes, the amp draw increase with slower gate drive is expected. Which photos?

I also question the component values connected to the 555 timer and to achieve the 3.7% duty cycle and 2.4KHz frequency. I'll look on the net for the component values - there are heaps of 555 calculators which will show the values for certain.

Yes that would be a good idea. Watch out for really low or really high values though. The spec sheet gives a recommended range of resistor and cap values for certain frequency ranges. I believe 100 or even 200 Ohm values are not within the recommended range.

I will consider a cheap analog scope. I've used them in electrical school college and they work quite well and fairly easy to use.

A decent scope is really worth having if you seriously want to understand circuits and learn to improve or modify how they work.

I have a concept for driving a resistor (inductor) I thought you might want to look at. Might be of use, maybe not. Or if you want to improve on it (eg to sense saturation using a differential op-amp mode). I hand drew it.
Imageshack - opampcmosinverterconcep

Andrew

I think there might be one or two small problems with your circuit, but in general I believe I get the idea. Your concept is good, but it might not behave quite as you might be hoping. Let me know if you'd like to work out the kinks and I'll try to help.

Regards,
.99

[Aaron]

Quantum spec circuit

I also question the component values connected to the 555 timer and to achieve the 3.7% duty cycle and 2.4KHz frequency. I'll look on the net for the component values - there are heaps of 555 calculators which will show the values for certain.

Ash,

Andrew can probably figure it out from this pic. Resistors from left to right 1.1k, 100ohm, 110ohm. I wrote specs on 2 caps, the cap on left he can lookup, I think it is 0.1uf but not absolutely sure.

Anyway, this setup will give a range that allows 3.7% duty and 2.4 kHz.




There are other configurations that will give those ranges but I know for sure this will do it as I was using this exact setup today for some tests.

[poynt99]

Probe Grounds

Quote from page 30 of the scope manual:

Ensure that all probe common leads are connected to the same voltage.

.99

[poynt99]

I don't think we need common grounds for this test as we are checking the current flow in each path at the same time. As long as your probe references are each isolated we should be ok. Otherwise the test becomes meaningless.

Cheers,

I know Aaron isn't reading my posts, and forgive me for stating the obvious, but shouldn't it be determined for certain if the probe grounds are separable or not? I'm not saying it isn't possible, but I've not seen a scope yet that has this capability. Not having the grounds separated electrically and attaching them to different places in the circuit could lead to some very undesirable effects.

.99

[gotoluc]

Hi everyone,

I've been busy with some temporary work (money making), like a squirrel stashing nuts before a long winter. So I have been short of time for research.

This post is to tell all of you that I do have the professionally made PC boards that Groundloop generously made and sent me free of charge of Rosemary's circuit and I still have 15 available. So don't be shy

Send your requests directly to my email: [email protected] with your full mailing address. Please note that I live in Canada, so delivery will take a little longer to the US or other countries.



Luc

[Harvey]

I have 4 shunts now.

neg batt to shunt to mosfet source
neg batt to shunt to neg rail for 555
pos batt to shunt to load
pos batt to shunt to pot meter for 555 power control

all are 0.25 ohms

I have scope channel 1 & 2 common ground and probe across each shunt on negative side.

If I use a common ground for the positive side, it will simply show the battery voltage. If it needs to be common ground to have channel 3 & 4 that means I can't do the negative and positive ones at the same time.

Also, I think it is much simpler to use one shunt to show the draw on the neg and one on the positive.

Anyway, I could log the data for the neg side then after that is done, I can log the data on the positive side. Comments?

I don't think we need common grounds for this test as we are checking the current flow in each path at the same time. As long as your probe references are each isolated we should be ok. Otherwise the test becomes meaningless.

Cheers,

[ashtweth]

Hi ALL.

Thanks so much Poynt/guys, just got in. Okay here is what i got back from Andrew. We can change the circuit and reproduce Harvey's/ Aaron's new if needed what ever it takes . Poynt got the cop circuit in the pdf and will get Andrew to get onto it my friend.

---------------------

Hey Poynt99,

I've measured the inductance of the resistor - comes out to about 20uH. The meter I have is not very accurate as its main function is a multi meter. I'll look at getting a proper LCR meter though.

I get the ripple during the on / off phase when I adjust the potentiometers to a certain setting.

The apparent duty cycle of 30% was the smallest I could get with the resistor / capacitor combination specified for that particular 555 timer circuit. I have observed that when you change the duty cycle on a 555 timer, it also changes the frequency. The scope I believe only has a bandwidth of about 10MHz.

I am using the 1N4007 flyback diode to charge into a capacitor, not to short out the spike.

I also observed that with the gate resistance set to 500 ohms, that the amp draw went up. The amp draw was minimum with no gate resistance (see photos).

I also question the component values connected to the 555 timer and to achieve the 3.7% duty cycle and 2.4KHz frequency. I'll look on the net for the component values - there are heaps of 555 calculators which will show the values for certain.

I will consider a cheap analog scope. I've used them in electrical school college and they work quite well and fairly easy to use.

I'll look more into the aperiodic resonance mode of mosfet operation.
The "Ainslie-Murakami Negative Dominant Waveform Generator" sounds interesting.

I have a concept for driving a resistor (inductor) I thought you might want to look at. Might be of use, maybe not. Or if you want to improve on it (eg to sense saturation using a differential op-amp mode). I hand drew it.
Imageshack - opampcmosinverterconcep

Andrew

[poynt99]

@.99,

The scope shots I am looking for: B(-) Mosfet Drain as Reference, B(+) Mosfet Drain as reference. Desired traces for each, Voltage, Current, Power.

The first one is simply an inverted Drain voltage measurement. The second one is simply the load voltage measurement (i.e. voltage across the load). I can do this with ease.

We've now established the voltage traces you want, but tell me precisely which devices you want current and power traces done for. Current has to be a device node, including batteries, and power is usually the device itself. If that's not making sense, you may want to elaborate on what you want so I can better understand the concept and get SPICE to plot it.

Also, an integration of those for a 1 minute run. I doubt that it is long enough for any thermal modeling to become active, but it is a start to where I am going with this.

1 minute is an eternity in SPICE when dealing with these frequencies. Usually it's not necessary to run that long anyway, which we'll get into when/if we get to that point.

Let's first see if we can accomplish task one, which is to get your equal ON and first clip times equal. Then we can move on to these measurements.

.99

[poynt99]

Thanks for the efforts here Poynt. Much obliged. There's some small fraction missing but I guess this efficiency is therefore at about 80%. Is that standard on a switching circuit?

Yes there is a small fraction of power dissipation missing from the 0.25 Ohm shunt, the 0.1 Ohm resistance in the battery, and a minute amount perhaps in the series 100 Ohm Gate resistor.

I excluded them because one, it's more work for me, and two, they were insignificant I thought, and one still gets the main picture I feel.

.99