DIYbio:Notebook/Open Gel Box 2.0/Power Supply

From OpenWetWare
Jump to navigationJump to search

Feb 25 2009 *User:Philip McIntosh 21:55 (CST):

Characteristics of prototype power supply built by Phil McIntosh:

  • 60-115VDC
  • Max current without cooling fan: 250mA@115VDC, 80mA@60VDC (estimated)
  • Max current with cooling fan: 300mA@115VDC (estimated), 110mA@60VDC
  • Includes onboard analog ammeter and digital voltmeter (not installed yet)
  • 0.6A fuse protection on output of input transformer
  • Based on Texas Instruments TL783 125VDC regulator (Figure 18 design in product data sheet)
  • Estimated cost of parts (without enclosure): $75.

Figure 5. TL783 Prototype on perf board

Figure 6. TL783 Prototype under bench test

Jan 31 2009 * 15:53, 31 January 2009 (EST):

Minimum Specs for open gel system 2.0's P/S.

  • Must be totally enclosed
  • Must work with 110/220 VAC input
  • Must work with 50Hz/60Hz AC input
  • Outputs two fixed voltages:
    • "FAST" setting is a fixed voltage anywhere in the range 90v-120v
    • "SLOW" setting is a fixed voltage anywhere in the range 50v-60v
    • These two voltages can be on a single connector (1 pair of banana plugs) or on separate connectors (2 separate pairs of banana plugs)
    • These output voltages do not need extensive filtering, as output ripple does not matter.
  • Must use red and black banana plug connectors
  • Must be capable of continuous 100 mA load
  • Must include either resettable CB protection or thermal protection; shutoff when electronics are over temperature or over voltage
  • Must have a 2nd DC peripheral supply, rated +12V to +18V @ 5W, to supply power to 4*LED illuminator units (each are 0.4W).
    • Output plug on the 2nd peripheral supply should be a medium-sized barrel connector with outside connection as ground.
  • May have a 3rd DC peripheral supply, rated +5V @ 5W, to supply generic logic circuits (like microcontroller boards). This supply could eliminate clutter when using several peripherals.
    • Output plug on the 3nd peripheral supply should be a different connector than the 2nd supply, to avoid mixing up the 2nd & 3rd supplies when connecting peripherals.

Jan 08, 2009, 10:34 PM

Specification Discussion

After talking to a number of people and reading the various discussion threads I propose the following specification for the Power Supply.

Edit each line to place a y (agree) or an n (disagree) for each of the line items. You do not need to justify any n's.

Feel free to add a line item you feel is important.

When all interested parties have had their "vote" we'll revisit and discuss any that have too many n's.

* 15:53, 31 January 2009 (EST): see new specs above
  • Must be totally enclosed (y)(y)
  • Must work with 110/220 VAC input (y)(y)
  • Must work with 50Hz/60Hz AC input (y)
  • Single channel 20-120 VDC output voltage, variable in 10 V steps (y)(y, at least)
  • Must use red and black banana plug connectors (y)(y)
  • Must have a voltmeter display (y)(y)
  • Must be capable of continuous 300 mA load (y)(y, at least)
TBD; current max requirement needs verification/calculation. Lower current would allow simplified design.
  • Must employ resettable CB protection (y)
  • Must include thermal protection; shutoff when electronics are over temperature
  • Should have a secondary +12V @ 300(?)mA output for connecting peripherals (fan, pump, etc)
CPU-size fan draws ~170mA. Pump draws (?) mA.

Jan 06, 2009

What we need: clean, stable, variable DC at 20V-100V, adjustable (possibly up to 600V).

Subsequent research shows the output doesn't need to be clean/stable. Ref. paper used half wave rectified AC circuit (which has high ripple output) which had same results as very-clean DC supply. Also, pulsed field electrophoresis even reverses polarity with beneficial effects, so even though we are not doing that technique here, stability isnt much of an issue either. * 02:52, 17 January 2009 (EST)

Consider: using a cheap computer ATX supply feeding a voltage regulator such as an LM317 feeding a DC/DC step-up built on the model introduced in (from ).

This EDN circuit "Build a transformerless 12V-to-180V dc/dc converter" at first looks great, though there is a big typo in "Table 2" which lists output current. They list output current in "mA" but it is actually in uA! So the "1000 mA" max is actually "1000 uA" == 1mA max. The text of the article references this. Based on the requirement of ~300mA for the gel, this particular circuit won't work, unfortunately. * 04:06, 15 January 2009 (EST):

For $10-$20 , you get plenty of overhead (i.e. 400W max neighborhood), 50/60Hz and 110/220V independence, amp draw regulation (probably? not sure what typical protection is in the low price range), and handy plug-in molex connectors. The last point is particularly salient, if you can power more than one device with a versatile power supply.

Low cost "brick" supply is available from manufacturers such as GlobTek (Taiwan); * 18:22, 8 January 2009 (EST):
For example, GT(M)21097CC SERIES 50W
Has Input Voltage 90v-260v @ 47Hz-63Hz (suitable for worldwide use)
Has Output Voltage +48.0 VDC @ 0 - 1.1 A
Cost is probably around $20-$30 per unit in low quantities.

I do not know how adding a voltage regulator (LM317 etc) and DC/DC step-ups after these affect their voltage regulating properties.

Converting a computer ATX power supply to a really useful lab power supply

Covers voltages under 24v. * 04:06, 15 January 2009 (EST):

How to Add Variable Voltage to Your ATX Based Bench Power Supply

Covers voltages under 24v. * 04:06, 15 January 2009 (EST):

Building a Power Supply