78005 is same as 7805 but with isolated tab. One of the options is 15v,so they probably bias the ground terminal by 1/3/7/10 volts to get 5/6/9/12/15. How you get the negative side without using a 7905 is beyond my paygrade
I *think* this data-sheet for the 78xx series adequately describes how you can get a negative [although they don’t appear to have actually implemented that].
Dispite these power supplies for the education market selling at a premium price the reason for using a positive 5 volt regulator for multiple voltages is probably cost. Not just component cost but the cost of stocking two parts rather than one and making sure they go in the righe place. I must admit the 78005 is new to me but again cost saving as there is no need for insulating washers if it has an isolated tab.
An engineer would have used a LM317 LM337 pair for this application. Much better performance all round.
As an aside griffin and others also sold fully variable unregulated low voltage AC / DC supplies. These are very handy and often appear at low cost. The smaller ones use an interesting dual winding low voltage output isolated variable transformer. The larger ones tend to have a standard mains variac and separate low voltage transformer.
On 19 September 2024 at 16:29Robert Atkinson 2 Said: […] I must admit the 78005 is new to me but again cost saving as there is no need for insulating washers if it has an isolated tab.
[…]
Strangely enough … the [hand-crafted from channel] heat-sinks are isolated from the case by the use of nylon screws and washers.
…. Is that because the Blue connection can be floating anywhere with respect to Earth ?? … Dunno !
When using a 78xx regulator at it’s designed voltage the tab is at 0V so if the chassis is also at 0V it does not need a insulator (no insulator improves thermal performance) when using it like this to produce higher voltages the Tab is at the output voltage minus 5V. Having checked a couple of datasheets I’m not convinced that the 78005 has an isolated tab. All the isolated 78xx regulators I’ve seen have had the tab totally encapsulated in the plastic.
There re of courase lots f ways to acheive the desired outcome but that power supply would not win any electronics design awards.
From what I can see of the wiring it is probably 2 identical supplies stacked one on top of the other. The +ve side of one goes to the +ve terminal and it’s 0v goes to the blue centre terminal. The other one has its +ve connected to the blue central terminal and its 0v to the 0v (black) terminal. Hence both can use the same 78005.
The nylon screws for the heatsink may be because the design originally use standard (non-isolated) 7805s.
the 78005 is ‘isolated’ and the picture looks as tho the tab is encased in plastic. Doesn’t Keith’s suggestion require 2 independant LT windings? Seems like a lot of fuss rather than a single centre tapped LT, one bridge rectifier and a 7805/7905 pair.
If you look at the photo of the inside, the two yellow wires from the transformer are separate from the two white wires that go to the top of the PCB. Each goes to its own back block (presumably the bridge rectifier) and then is wired to its own capacitor.
I agree it’s a lot of fuss for such a simple supply. Maybe its a really old design that has continued to be built – I seem to recall when the 7805 first came out (and seemed like magic) the 7905 didn’t exist, it came a bit later.
If you look at the photo of the inside, the two yellow wires from the transformer are separate from the two white wires that go to the top of the PCB. Each goes to its own back block (presumably the bridge rectifier) and then is wired to its own capacitor.
[…]
Indeed … it appears to be just two simple and identical circuits, rather expensively buit.
The Toshiba device Duncan linked to is a 780xxAP. It is the P (plastic) suffix that makes this device isolated. It is totaly encapsulated in plastic. The devices in Michael’s power supply are neither Toshiba or encapsulated.
I agree with Keith, it appears to be two positive supples stacked. Quite strange. Also strange is the stack of presets for the different voltages. This could easily be done with one less bridge rectifier, LM317T, LM337T and fixed resistors.
The LM317 is much more rugged and being designed for the application can use fixed resistors to set the voltage. With a 78xxx the resistor value rquired will depend on the quiesent current of the particular device so you need an adjustment. The LM317 has about 25 transistors and 4 zeners compared to 17 and one for the 78xxx.
It contains no schematics or explanation, but describes some performance tests on a selection of Stabilised Power Supplies being offered for use in Scottish schools.
The Griffin unit received this damning assessment:
I’m not surprised.
Mind you there is a symbol (dashes under a solid line) next to “volts” above the selector switch that indicates the output may be pulsating DC.
On 19 September 2024 at 16:29Robert Atkinson 2 Said:
[…] I must admit the 78005 is new to me but again cost saving as there is no need for insulating washers if it has an isolated tab.
[…]
Strangely enough … the [hand-crafted from channel] heat-sinks are isolated from the case by the use of nylon screws and washers.
…. Is that because the Blue connection can be floating anywhere with respect to Earth ?? … Dunno !
MichaelG.
This could well be part of the shenanigans needed to get a positive 78xx to regulate a negative rail. The basic circuit is shown in the TI datasheet without comment. I guess because in most circumstances a negative regulator is the straightforward answer, and better! Robert is likely correct in that it was done in the Griffin PSU for cost reasons, eliminating one component from inventory.
I only do basic electronic design. Too keep design simple, it’s easier for me to think of power rails being one way round. But though conventional, it’s not strictly necessary, because the electronics don’t know what ‘common’ is. Provided the circuit is arranged so that components stay within their ratings all will be well.
In the example above, the positive output of the 78xx is grounded, whilst the negative rail is allowed to float. Although it works, the designer must provide a floating power supply, and sort out all the downstream consequences that has on the rest of the equipment. For example, in a conventional circuit, the 78xx tab is at ground potential and can be safely bolted direct to an earthed heat-sink. In the negative circuit above, the 78xx tab is at -xx volts, making it essential to electrically isolate the heat-sink. Having to isolate the heat-sink could easily be more expensive than keeping the design simple by coughing up for a 79xx negative regulator. We don’t know if the designer did a good job or not at the time.
Another example comes from the world of Operational Amplifiers. These remarkable devices operate on a split power rail, say -12v, 0, and +12v. This is inconvenient compared with a single rail power supply, so it’s not unusual to find Op Amp circuits arranged around an artificial mid-point where -12, 0, and +12v is replaced by 0, +12 and +24. What normally used 0V as centre, now runs at 12V, and the negative rail is ground. The chip can’t tell the difference.
Just shows how careful you have to be. The Toshiba device I linked has exactly the same part number as the one in Michael’s photo, but Michael’s doesn’t have the plastic covered tab.
OP amps are now available which claim to work right down to the ground rail. The problem with earlier ones was that if supplied from a single rail supply they gave up the ghost on small input voltages. In one application where I had this problem I added a little chip which generated a voltage opposite to the supply voltage.
Author
Posts
Viewing 21 posts - 1 through 21 (of 21 total)
Please log in to reply to this topic. Registering is free and easy using the links on the menu at the top of this page.