MAT-04/CA3046 Characterization

[Tim Ressel]

Martin,

Yes, I agree it would be dangerous to assume this kind of thing. That is why I
did the characterization run. I did not show the gruesome details, but here they
are:

Temp		MAT-04
Chamber	Unit A	Unit B	Unit C	Unit D
39.9		.5451		.5450		.5451		.5451
45.0		.5338		.5337		.5338		.5338
49.9		.5227		.5226		.5227		.5227
55.0		.5111		.5110		.5111		.5112
60.0		.4999		.4998		.4999		.4999

As you can see, the matching across the die is quite good. The max deviation is
0.0002V, which at a slope of -2.26mV/Deg C, comes out to a variance of 0.452 C
across the transistors. Also keep in mind that the thermal resistance of the
plastic case is 39 C/watt, which means the bulk of the heat is kept inside the
case. This will help to keep the die iso-thermal.

To address your other concern: yes, the temperature operating range is an issue.
I am in contact with Analog Devices to determine what the power constraints are
on the trannies. I am at the current limit, but if I can boost the voltage, then
I will get more power e.g. more heating.

I am pushing this design because the MAT-04 will make a superior expo converter
to the CA3046 while keeping the built-in heater feature. I find this an
attractive design solution, but I need to know if it really works. Its the
engineer in me.

Stay tuned.

Tim Ressel--Compliance Engineer
Hewlett-Packard
Verifone Division
916-630-2541  
tim_r1 at verifone.com                     



-----Original Message-----
From: Martin Czech [mailto:czech at Micronas.Com]
Sent: Thursday, June 08, 2000 12:34 AM
To: Tim_R1 at verifone.com
Cc: synth-diy at node12b53.a2000.nl
Subject: Re: MAT-04/CA3046 Characterization


I think it is dangerous to rely on transistor matching through
different places from a wafer, or even lots.
I.e. the voltage versus temperature could be different for each individual
transistor. This means that the servo loop must be calibrated for each 
individual sensor, otherwise the chip temperature is too low or too hot.


m.c.

:::From: Tim Ressel <Tim_R1 at verifone.com>
:::To: synth-diy at node12b53.a2000.nl
:::Subject: MAT-04/CA3046 Characterization
:::Date: Tue, 6 Jun 2000 16:07:22 -0700 
:::
:::Yo,
:::
:::I just completed the characterization run. Here are the results:
:::
:::Chamber	MAT-04	MAT-04	CA3046
:::Temp		Temp		Voltage	Voltage
:::39.9		39.9		.5451		.6892
:::45.0		44.8		.5338		.6803
:::49.9		49.6		.5227		.6715
:::55.0		54.7		.5111		.6624
:::60.0		59.5		.4999		.6535
:::
:::PS Voltage: 9.98V
:::Bias resistor: 10.0K 1%
:::
:::As you can see, the MAT-04 voltage is way different than the CA3046. This 
info
:::is useful for setting the bridge voltage on the AN-299 heater scheme. For
:::example, if you wanted the MAT-04 to run at 40C, you would select a voltage
:::divider pair to yield .5451 volts, or close to it at least. The stock circuit
:::yielded 0.6273V. 
:::
:::Next I am going to run the AN299 circuit with the MAT-04 over temperature to 
see
:::how it does. Unfortunately, when I rigged the circuit in the chamber it 
didn't
:::work. I somehow blew the MAT-04. The spare is working okay. After I glue the
:::thermocouple to the back of the new MAT-04 I will begin the next run.
:::
:::Tim Ressel--Compliance Engineer
:::Hewlett-Packard
:::Verifone Division
:::916-630-2541  
:::tim_r1 at verifone.com