ArcMania Microconverter performance data

[Doug S]

I recently purchased one of the ArcMania microconverter boards from Wayne's Sandwich Shoppe. The board is nicely laid out and the construction quality is high. It is hard to imagine making a convertor of this type any smaller without compromising performance or layout quality. Here is performance data using a very low Vf 1W luxeon as a load.

<ul type="square">
Vin(V)**Iin(mA)**Vout(V)**Iout(mA)**Eff(%)
3.02------357-------3.14-------298--------87.6
2.00------365-------3.02-------186--------77.7
1.503-----380-------2.97-------137--------72.0
1.305-----380-------2.93-------114--------68.2
1.199-----376-------2.91-------102--------66.6
1.100-----363-------2.89-------087--------63.8
1.001-----308-------2.84-------061--------57.0
0.897-----158-------2.77-------030--------59.4
0.875-----103-------2.74-------021--------64.6
0.85 shuts off
[/list] Operating frequency ranged from 360kHz at 2.00V to 140kHz at 1.001V. Those checking my math from the raw data above will find that my efficiency data is 0.8% higher than computed. The efficiency column contains a 0.8% correction resulting from cross calibration of the various instruments used.
Notice the very flat input current from 1.1V to 3.0V. This is a Zetex 300 design. It incorporates a trick to flatten the input current vs voltage characteristic. The Isense pin is biased by a divided fraction of the input voltage. This results in a flatter Iin vs Vin curve than a dumber converter such as used in the Dorcy 1AAA or the ARC 1AAA.
For comparison, here is data for the Dorcy 1AAA converter powering a considerably higher Vf luxeon:
<ul type="square">
Vin(V) * Iin(mA) * Vout(V) * Iout(mA) * Eff(%)
0.503----41.1-------2.74--------5.7-------74.7
.799-----74.2-------2.86-------15.7-------75.9
.998-----112--------2.93-------28.8-------74.7
1.106----212--------3.05-------57.7-------74.3
1.196----295--------3.11-------81.1-------70.7
1.300----392--------3.17------108.5-------66.7
1.406----486--------3.19------137.6-------63.4
1.500----536--------3.20------157.5-------61.9
[/list]
Notice that the Iout of the Dorcy varies much more in the 1.1V to 1.5V range. The Dorcy does continue to operate down to much lower voltages and its efficiency is better at the lower end of the Vin range.
A couple of ARC 1AAA circuits are on the way to me and I will post equivalent data for the ARC circuit when available.

 

[Chief_Wiggum]

Doug,

Thanks for the information! I just purchased one of these myself and was curious to see how it performed.

I noticed you ran it all the way up to 3V. ArcMania told me he wasn't sure whether or not the transistor could handle the voltage. In your opinion can this board be used in a 2 AAA configuration?

 

[Doug S]

[ QUOTE ]
Chief_Wiggum said:
Doug,

I noticed you ran it all the way up to 3V. ArcMania told me he wasn't sure whether or not the transistor could handle the voltage. In your opinion can this board be used in a 2 AAA configuration?



[/ QUOTE ]
Yes. Looking at the efficiency data, the board is actually dissapating more power at a Vin of 1.5V than it is at 3V.

 

[Chief_Wiggum]

Thank you, Doug. 1 penlight coming up /ubbthreads/images/graemlins/smile.gif

 

[Doug Owen]

Well done (other) Doug!

Thanks for sharing such useful data.

Doug Owen

 

[Doug S]

[ QUOTE ]
Doug Owen said:
Well done (other) Doug!



[/ QUOTE ]

Thanks Doug
-Doug

Chief: Now that I have slept on your question, I have a bit to add. This design biases the Isense input to the ZXSC300 with the input voltage reduced by a 10 ohm/1740 ohm voltage divider. The Isense of the IC has a typical Isense threshold of 19mV. This should allow the circuit to operate up to an input of (.019V)(1750/10)=3.32V. Unfortunately, the datasheet shows a large variation of possible Isense with a min/max of 14mV to 24mV. If you were unfortunate and got one with the min value of 14mV, the circuit would not turn on for voltages higher than (.014V)(1750/10)=2.45V. If this is the case and you want to operate at Vin=3V you could either increase the value of the 1740 ohm resistor a bit (best) or remove it from the board. The resistors used for the voltage divider are physically *very* small, 0201 case size I believe. BTW, the actual Rsense resistance in formed by a length of narrow trace on the PCB. On mine it measures 32 milliohm.

 

[Chief_Wiggum]

Doug, thanks again for the further info. After looking at the board, there is _no_ way I can remove or change the value of that resistor. It's far too small.

If I bench test it at 3v, and it doesn't turn on, then I'll know what the problem is. My only question is can any damage be done by testing it this way?

 

[Mark_Larson]

Doug, where did you buy that board? I can't find it on the Sandwich Shoppe. /ubbthreads/images/graemlins/confused.gif

 

[Doug S]

[ QUOTE ]
Chief_Wiggum said:
If I bench test it at 3v, and it doesn't turn on, then I'll know what the problem is. My only question is can any damage be done by testing it this way?



[/ QUOTE ]
No, provided that the applied polarity is correct and an LED is connected as a load. Double check before applying power!

 

[Doug S]

[ QUOTE ]
Mark_Larson said:
Doug, where did you buy that board? I can't find it on the Sandwich Shoppe. /ubbthreads/images/graemlins/confused.gif

[/ QUOTE ]
At the Sandwich Shoppe, click on MJ converters [or maybe it is MJ products] on the left side of the page. $27.

 

[Doug S]

As promised, equivalent test data for the ARC 1AAA circuit has been posted. I have placed it in a new thread here:
ARC 1AAA circuit data