BeebSID

for bbc micro/electron hardware, peripherals & programming issues (NOT emulators!)
User avatar
MartinB
Posts: 5273
Joined: Mon Mar 31, 2008 9:04 pm
Location: Obscurity
Contact:

Re: BeebSID

Post by MartinB » Fri Oct 16, 2009 11:10 am

No worries and I take your point about the gobbledygook :wink:

I'm re-working the drawing and will then start down the path of getting some PCB's made. I think that will be the best way forward because just making one on stripboard is rather time-consuming and I should know!

A "circle of SID chips" eh? Now you've said that publicly I would advise that you immediately lock your windows and doors :lol:

Martin

User avatar
MartinB
Posts: 5273
Joined: Mon Mar 31, 2008 9:04 pm
Location: Obscurity
Contact:

Re: BeebSID

Post by MartinB » Mon Oct 19, 2009 9:30 pm

A guy by the name of Peter Rendle has posted on the Mailing List tonight and is wanting to dispose of some old computer kit. Amongst other things he has two broken C64's. Anyone near to Surrey/Sussex could grab them and harvest the SID's.... :wink:

The revised schematic is well under way but slower than my usual sketches because I now have to do it electronically and in a package that can then turn it into a PCB layout. I will publish this schematic because it can still be built on Verobard or Eurocard if anyone fancies that route.

If you are wondering how long it will take me, I'm new to this and the first time I tested the auto-routing for the layout the confirmation box did not say "Are you sure Y/N ?", it said "You cannot be serious Y/N ?" :roll:

Martin

Prime
Posts: 2837
Joined: Sun May 31, 2009 11:52 pm
Contact:

Re: BeebSID

Post by Prime » Mon Oct 19, 2009 10:08 pm

MartinB wrote:A guy by the name of Peter Rendle has posted on the Mailing
If you are wondering how long it will take me, I'm new to this and the first time I tested the auto-routing for the layout the confirmation box did not say "Are you sure Y/N ?", it said "You cannot be serious Y/N ?" :roll:
Personally I don't tend to auto-route as I've found that you end up with a board that is sub-optimal :( Especially if you are making the board by hand you want to minimize the number of layer changes etc.

Mind I really only have experience of the auto-router in Eagle 4.x, not tried any other auto-routers, or for that matter the one in 5.x :)

Cheers.

Phill.

User avatar
MartinB
Posts: 5273
Joined: Mon Mar 31, 2008 9:04 pm
Location: Obscurity
Contact:

Re: BeebSID

Post by MartinB » Mon Oct 19, 2009 10:18 pm

Thanks Phill and yeah, I think I'd already come to that conclusion but for a simple layout like this, a dummy auto-routing at least tends to show if what you're asking is physically possible - novice remember :wink:


Edit : Just read what you said again...
Prime wrote:Especially if you are making the board by hand...
By what now? :shock: :lol:

Martin

User avatar
MartinB
Posts: 5273
Joined: Mon Mar 31, 2008 9:04 pm
Location: Obscurity
Contact:

Re: BeebSID

Post by MartinB » Tue Oct 27, 2009 10:15 pm

Don't get too excited iomanoid, this is just a work in progress update but you can of course print off the pics and throw them into the circle of SIDs with some eye_of_newt and intensify the chanting... :D
schematic no analogue_s.JPG
schematic no analogue_s.JPG (23.86 KiB) Viewed 3380 times
pcb no analogue.JPG
pcb no analogue.JPG (55.68 KiB) Viewed 3381 times
The remaing analogue discretes are missing but some general questions for Phill, Mark, Alan (etc.?) - what's the best way of deciding where to physically place components? Do you just keep moving and routing until it's ergonomically sound (and works) or are there some clever tricks? Does that example layout look way too complex for the relatively simple schematic? (I'm using DipTrace btw because I have flirted with it years ago and that layout is my part placing followed by auto-route.)

Any general advice is welcome... :D

Martin

User avatar
retroclinic
Posts: 3034
Joined: Thu Jul 03, 2008 1:22 pm
Location: East Riding of Yorkshire
Contact:

Re: BeebSID

Post by retroclinic » Tue Oct 27, 2009 10:41 pm

Hi.

Looks good.

I would put the pass thru and the main connector next to each other, simplifies the routing between them, or ditch it all together, you can always do a pass thru using a plug on the ribbon cable. Add the option for termination resistor packs, although you may not need to fit them.

No caps (and nowhere to plug power in?) - put at least one electrolytic, 22uF ish, and a decoupling ceramic 0.1 for each chip. Also, add an LED - they look 8)

Expand the board, and put 3mm mounting holes at each corner, so you can use sticky pads with posts to hold the board in a case.

Give an audio out on a phono (or a header), and put a jumper to allow the audio to be sent back thru the 1MHz cable, if required.

I tend to route mostly manually - or I do now - I use to use autorouting, but since I did DataCentre I found I can get a neater layout by doing most of it by hand, and just letting the autorouter "suggest" final routes for hard to track signals. (As well as Eagle getting it's e-nickers in a total knot when I tried to autoroute the entire board!).

Mark.
Image

User avatar
MartinB
Posts: 5273
Joined: Mon Mar 31, 2008 9:04 pm
Location: Obscurity
Contact:

Re: BeebSID

Post by MartinB » Tue Oct 27, 2009 10:56 pm

Yeah, I have the caps covered including decouplers, they are all there on Tom's prototype, just not on that schematic yet. I'm going to use two lo-watt regs so there will be at least a couple of electrolytics to cater for 'less than quality' dc power adaptors and yes, there will be a DC power jack. Phono yes, jumper yes - all designed in already. Flippin 'eck, I did say that this is a work in progress update :lol:

I have got layouts with the two pass-thru's together but I think I'll need to manually route those because DT didn't seem to see the obvious :roll:. I was just playing around with layouts and thought I'd ask if there was some exact science that I was missing :?

One more then - how does one decide the default lead spacing for discretes such as caps? For example, there will be the two 6800pf polystyrene filter caps - are axial types the same length for all manufacturers or is there some standard default?

Still, the good news is that you didn't employ the dreaded shaky finger :D

Martin

User avatar
retroclinic
Posts: 3034
Joined: Thu Jul 03, 2008 1:22 pm
Location: East Riding of Yorkshire
Contact:

Re: BeebSID

Post by retroclinic » Tue Oct 27, 2009 11:09 pm

Hehe, yes I realised it was "in progress", just wanted to make all the suggestions up front just incase.

As for spacing for caps, what I do is decide on a component, and either order some samples or do the board as per the spec of the component. It's always a good idea to have the components infront of you when you do the layout, as you can (well, Eagle can) print out an exact size copy of the board, and make sure it all fits on paper before committing it to copper.

Mark.
Image

User avatar
MartinB
Posts: 5273
Joined: Mon Mar 31, 2008 9:04 pm
Location: Obscurity
Contact:

Re: BeebSID

Post by MartinB » Tue Oct 27, 2009 11:38 pm

Me kidding too - your input is most welcome :D

The last time I actually committed something to PCB was waaaay back when and that was a 32k DRAM extension board for for my Ohio Superboard. Due to a delightful mixture of PCB inexperience and pure hamfistedness, that ended up as a 16k-by-6-Bit expansion board #-o

I will undoubtedly be back..... :wink:

Martin

User avatar
AlanD
Posts: 241
Joined: Fri Jan 09, 2009 8:30 pm
Contact:

Re: BeebSID

Post by AlanD » Wed Oct 28, 2009 6:18 pm

Hello Martin

i agree with Mark about the pass thru put them next to each other

personally i never auto route boards they always finishes up bigger then a hand routed board and are rarley as pleasing to look at.

With your SID chip in mind i would avoid running any of the 1MHz bus signals under the SID chip they could cause pick up .In fact i would place the SID chip so you can approach one side of it with the digital signals (ie keep all the Digital in one area and the analogue in another area). Also the Logic chips i would have a play around with swapping about the gates, almost certainly by swapping around the inputs to the 8 input nand the tracking be simplified

Also with double sided boards i try to track as much as possible on the top layer this allows me to get maximum pour coverage for a Ground on the bottom layer.

I always add
headers for connectors
a large GND pad (to solder a GND wire to incase you have to debug the board)

Add a power indicator

just because you track in these extra bits does not mean you have to fit them to all boards but they may prove very useful.

when i intend to box something i will usually try to get the box before hand so i can tailor the board to fit ie mounting holes and board outline

dont forget to either silk screen or copper your name and details everyone should know who the clever person behind the design is!!

The rest is just time for tinkering with the routing.

just some of my random thoughts hope they help

Aland

User avatar
MartinB
Posts: 5273
Joined: Mon Mar 31, 2008 9:04 pm
Location: Obscurity
Contact:

Re: BeebSID

Post by MartinB » Wed Oct 28, 2009 6:42 pm

That's brilliant thanks Alan - just the sort of tips I'm looking for :D. Even if you're into electronics, PCB design can still be a bit of a mystery until you've had some experience. I know it's only a simple project but it's plenty for a beginner like me :)

Thanks again, Martin

User avatar
AlanD
Posts: 241
Joined: Fri Jan 09, 2009 8:30 pm
Contact:

Re: BeebSID

Post by AlanD » Wed Oct 28, 2009 6:58 pm

Hello Martin

one more thing i usually add in the copper is the following text

top layer (on the top layer obviously)
bottom layer (on the bottom layer obviously)
topbottom.JPG
topbottom.JPG (98.14 KiB) Viewed 3253 times

this removes any ambiguity as to which layer goes where for for board manufacturer

AlanD
Last edited by AlanD on Wed Oct 28, 2009 7:11 pm, edited 2 times in total.

Prime
Posts: 2837
Joined: Sun May 31, 2009 11:52 pm
Contact:

Re: BeebSID

Post by Prime » Wed Oct 28, 2009 7:00 pm

The other thing that might be worth doing is having some jumpers/switches so that you can select either +12V or +9V to the SID, that way you can fit a 9Vreg
and feed the board with +12V e.g. from the beeb PSU, and then select at build time which version of the SID you have. Also usefull if you need to change it later.

Cheers.

Phill.

User avatar
MartinB
Posts: 5273
Joined: Mon Mar 31, 2008 9:04 pm
Location: Obscurity
Contact:

Re: BeebSID

Post by MartinB » Wed Oct 28, 2009 7:39 pm

Ah - is there a version of the SID with a +12v Vdd then?

I planned to have a nominal (minimum) DC input of +12v and then use two small regulators - either 78L05/09 if the currents are small (to avoid a heatsink) or the ordinary 1A 7805/09 which have more than sufficient sink in the tab.

But yes, if there's such a thing as a 5v/12v SID then a jumper won't be a problem (apart from the fact I'm now getting a headache of course... :wink:)

@ Alan : Top & Bottom layer marking sounds good :D

Martin

User avatar
SarahWalker
Posts: 1220
Joined: Fri Jan 14, 2005 3:56 pm
Contact:

Re: BeebSID

Post by SarahWalker » Wed Oct 28, 2009 8:05 pm

6581 SID is 12V. Unfortunately it also needs different filter caps - 1000pF instead of 6800pF (at least that's what the datasheet says). There's also a 6582, though that's 9V/6800pF.

User avatar
MartinB
Posts: 5273
Joined: Mon Mar 31, 2008 9:04 pm
Location: Obscurity
Contact:

Re: BeebSID

Post by MartinB » Wed Oct 28, 2009 9:20 pm

Cheers Tom :). Ok, a jumper for 12v direct or 9v on-board regulated it is then. I will have to stipulate though that if a 12v SID is used, the 12v DC input must be truly regulated, i.e. from the Beeb or from a properly regulated DC adaptor. Some adaptors, although looking the same, are not actually regulated and push out much larger voltages unless significantly on load. These types would not be good for SID's health [-X. Yes, I could fit a zener but I have to draw the line somewhere and it's already occurred to me that having allowed for DC adaptors I'll probably have to fit a diode for reverse polarity protection :roll:

The different (and probably physically smaller) filter caps won't be a problem because as long as they're axial types then it just means that they'll need slightly longer leads before the bend.

Martin

User avatar
retroclinic
Posts: 3034
Joined: Thu Jul 03, 2008 1:22 pm
Location: East Riding of Yorkshire
Contact:

Re: BeebSID

Post by retroclinic » Wed Oct 28, 2009 9:46 pm

MartinB wrote:I will have to stipulate though that if a 12v SID is used, the 12v DC input must be truly regulated, i.e. from the Beeb or from a properly regulated DC adaptor.
Sorry, to put a spanner in the works, but the 12v on the Beeb's PSU is not regulated either. It varies depending on the load of the 5v. A Zener would be a good idea as you say if you want it Beeb powered.

Mark.
Image

User avatar
MartinB
Posts: 5273
Joined: Mon Mar 31, 2008 9:04 pm
Location: Obscurity
Contact:

Re: BeebSID

Post by MartinB » Wed Oct 28, 2009 9:51 pm

Really? As in... what sort of range unregulated?

I've certainly never noticed, it's always as near as matters 12v whenever I've measured it on a running Beeb :?

Martin

User avatar
AlanD
Posts: 241
Joined: Fri Jan 09, 2009 8:30 pm
Contact:

Re: BeebSID

Post by AlanD » Wed Oct 28, 2009 10:09 pm

Hello Martin

i would not use a diode for polarity protection i would use a MOSFET like this almost no volt drop if you do it like this instead of the 0.6 or so from a silicon diode
mosfet.JPG
mosfet.JPG (60.16 KiB) Viewed 3178 times

AlanD

User avatar
MartinB
Posts: 5273
Joined: Mon Mar 31, 2008 9:04 pm
Location: Obscurity
Contact:

Re: BeebSID

Post by MartinB » Thu Oct 29, 2009 7:37 pm

@ Alan : Yes, of course you're right about the 0.6v drop but the tolerance of the SID and/or the regulator would be ok with that. I must confess to be deliberately treading a 'safe' line because using a device such as the one you suggest then puts me in a flat spin again regarding package types (pin patterns) for the PCB :roll: and, I was hoping that at the end of the day, people could take a PCB and a SID and pretty much get everything else from easy-shop places such as Maplin. I am not intending to build these myself and want to, as far as possible, keep the kit list right down at the enthusiastic hobbyists bits & pieces draw level :wink:

Anyway, I'm already thinking of renaming it to BeebSOD...

@ Mark : You didn't expand on that alleged 12v spanner? I don't have a detailed circuit diagram of the PSU but the simple schematic suggests it is regulated but as function of the 5v regulation - hence regulated. My hands-on tests also show it's a damn good 12v from off-load to the max. rated loads of the PSU :D

(...and that's what Tom's prototype BeebSID uses so just stop all this smart-ass scare-mongering :lol:)

Martin

User avatar
retroclinic
Posts: 3034
Joined: Thu Jul 03, 2008 1:22 pm
Location: East Riding of Yorkshire
Contact:

Re: BeebSID

Post by retroclinic » Thu Oct 29, 2009 8:02 pm

Hi Martin.

Take a look at the schem for the UK Aztec PSU
PSU Schem.jpg
PSU Schem.jpg (56.14 KiB) Viewed 3077 times
The regulation is fed from the +5v line. The 12v is produced by feeding the 5v winding back thru the transformer via a rectifier, the out of which is fed straight out via a smoothing cap and a coil. The adjust resistors adjust the 5v AND the 12v, because the 12v does not have it's own independant regulation.

What this all means is that if the load on the 5v goes up, the regulation will up the PWM width to compensate and keep the 5v at 5v. This will increase the voltage on the unregulated outputs - the 5v part will be compensated for but not the winding that produces the further 7v. Also if a load is placed on the 12v, the voltage will go down, but it will not be compensated for as it is not the line that is regulated.

I've found power supplies ranging from 11.2 to 13.3v on the 12v line with a fixed 1A load on the 5v, which is fine for the motor drive of a disk drive. What sort of regulation does the SID need? If it's in that range, then there shouldn't be a problem.

Mark.
Image

User avatar
MartinB
Posts: 5273
Joined: Mon Mar 31, 2008 9:04 pm
Location: Obscurity
Contact:

Re: BeebSID

Post by MartinB » Thu Oct 29, 2009 8:17 pm

The 9v or 12v Vdd for the SID is not a straight power supply as such, it is for the analogue circuitry and it will work from quite a wide range. Commodore quote +-5% to guarantee their analogue audio specs but a wider range will just impose envelope constaints. Max. Vdd is +14v.

I'm just splitting hairs but 11.2 to 13.3 is not the sort of 'unregulated' I was worried about - some of these adaptors start at 15 or 16 volts which is not good!

Thanks for the explanation - very comprehensive :D

Martin

User avatar
MartinB
Posts: 5273
Joined: Mon Mar 31, 2008 9:04 pm
Location: Obscurity
Contact:

Re: BeebSID

Post by MartinB » Thu Nov 12, 2009 12:38 pm

Ok people - here's my proposed full schematic for a 'production' BeebSID pasted as quickview and attached for scrutiny. Naturally, I have some questions.... :wink:
BeebSID_s.JPG
BeebSID_s.JPG (27.69 KiB) Viewed 2917 times
I think all the discussed 'features' have been included. Worthy of note perhaps are...

* LK1 allows the SID audio out to be fed back into the Beeb internal audio amp.
* LK2 allows the on-board 9v regulator to be bypassed such that the incoming DC supply can either be a very good 9v/12v supply (for 9v or 12v SIDs) or can be a good or shitty 12v or greater for 9v SIDs only.
* DC input is via a familiar barrel type connector - size TBD
* Audio output is via an RCA/Phono socket.
* Red LED (Yay!) to indicate the presence of +5v

@ AlanD : I decided I would go with your suggestion of a P-MOSFET for reverse polarity protection but, not being terribly familiar with this usage, I'm slightly confused about what type to specify and/or to pad for on the PCB. The device you suggested is (I think?) a 7A device :shock: and presumably is a TO220 package? I was looking for a smaller, hopefully say TO92, device but I'm not sure what parameter will define it's suitability in the application we are using. Is it I-Drain or I-Drain(On) :?. I picked a device that seems to be more readily available than the NDT456 (and it's in the DipTrace library :wink:) but it shows an Id of only -160mA and an Id(on) of -400mA. The latter would be adequate but the former wouldn't. I'd like to avoid a third TO220 device if possible (78L0x weren't up to the job :() so some advice please.... :)

(For others who are confused, the MOSFET can be omitted with a link if you are happy that you won't reverse connect the supply or, provided your supplies are not marginal, can be replaced with a simple diode. More details later.)

@ Anyone : The choice of support caps for 780x applications seems to be a completely random science with conflicting data everwhere you look. I normally just select by intuition so have just gone for a 47 across the common input and a 10 across each output. Feels ok but any thoughts with explanations welcome....

I'm on with the routing but being a PCB novice on a steep learning curve, nothing is happening quickly apart from the rate at which I get headaches #-o

Martin

*** EDIT : Schematic superceded, see further down. Please re-download from there ***
Attachments
BeebSID.zip
(214.24 KiB) Downloaded 95 times
Last edited by MartinB on Thu Nov 12, 2009 9:47 pm, edited 2 times in total.

User avatar
retroclinic
Posts: 3034
Joined: Thu Jul 03, 2008 1:22 pm
Location: East Riding of Yorkshire
Contact:

Re: BeebSID

Post by retroclinic » Thu Nov 12, 2009 2:23 pm

MartinB wrote:The choice of support caps for 780x applications seems to be a completely random science with conflicting data everwhere you look. I normally just select by intuition so have just gone for a 47 across the common input and a 10 across each output. Feels ok but any thoughts with explanations welcome....
Looking good - except for the 47uF cap - put the +ve after the FET, otherwise reversing the power supply will cause the cap to explode in a cloud of yellow goo.

About the caps. Values are dependant on load, and the ones you've selected will do fine. The important thing to remember is the value on the input must be greater than the value on the output. This is because if the supply fails, and the input cap value is smaller, it will discharge before the cap on the output, causing a reverse polarity in the regulator, which may do nothing, but could also cause the regulator to fail (that's what I understood from Datasheets and doing PSU design many moons ago).

Mark.
Image

User avatar
MartinB
Posts: 5273
Joined: Mon Mar 31, 2008 9:04 pm
Location: Obscurity
Contact:

Re: BeebSID

Post by MartinB » Thu Nov 12, 2009 5:21 pm

Ah yes, cheers Mark on the FET/Cap order :D - I kept juggling the front end components but decided in the end that protection is necessary since a lot of these adaptors have reversible plugs to cater for all possible kit. Still, it wouldn't be a proper Beeb accessory if there wasn't the potential for at least one exploding capacitor :wink:

Thanks for the info on the regulator caps - the manufacturers also often suggest a 330n on the input which seems to be concerned with the possibilty of oscillation in the presence of severe ripple but I think one could get carried away so I'll just stick with what's always worked.

I'll wait for Alan's thoughts on the MOSFET and then we're away :D

Martin

User avatar
MartinB
Posts: 5273
Joined: Mon Mar 31, 2008 9:04 pm
Location: Obscurity
Contact:

Re: BeebSID

Post by MartinB » Thu Nov 12, 2009 9:39 pm

And again with the input cap moved to the protected side of Q1 and the power supply layout improved for clarity...
BeebSID_s.JPG
BeebSID_s.JPG (28.15 KiB) Viewed 2864 times
The circuit shown is now complete and can be built by any industrious hardware types. I am progressing the PCB but I'm not sure how long it'll take or how much they'll eventually cost. When I've got a submittable design, I'll conduct a straw poll to see how many might be required and then get some quotes.

If you want to build your own on Vero or Euro then feel free - the following notes may help :

* The chip gates (and therefore pin numbers) can be switched around as required to ease the layout provided the logic is not altered.

* On the power side, you can dump both regulators if you have a 12v SID and are powering the board from the Beeb disc drive connector but you must then have two rails coming in, +5v to Vcc and +12v to Vdd. If you just have a good, stable 12v supply and a 12v SID then you can lose the 7809 regulator and set LK2 South (schematic ref)

* Q1 is present for reverse polarity protection but can be left out altogether by adding a shorting link from D to S or can be replaced by a simple silicon diode (eg 1N4001) with anode to D and cathode to S. If using a diode and the 7809, your supply voltage must not be less than about 11.6v because the volt drop across the diode may cause the regulator output to fall away or oscillate.

* Electrolytic caps should be 25v working, the 6800pf filter caps are Polyester and the 0.1 decoupling caps should be mounted as close as possible to each chip's supply pin.

* The 1MHz pass-through connector can be lost, as can both if you want to solder the ribbon directly to the board. The latter is a pain as I know from the prototype :roll:

I'll add an edit to the previous post to warn that the first posted schematic has been superceded and this is now the definitive design ( [-o< ) apart perhaps from the precise type of Q1.

Martin

*** EDIT : Schematic superceded again, see further down. Please re-download from there ***
Attachments
BeebSID.zip
(223.9 KiB) Downloaded 94 times
Last edited by MartinB on Thu Nov 12, 2009 11:48 pm, edited 1 time in total.

User avatar
IanS
Posts: 1142
Joined: Mon Aug 31, 2009 6:02 pm
Contact:

Re: BeebSID

Post by IanS » Thu Nov 12, 2009 10:57 pm

IC2.2 pin 6 & IC2.3 pin 9, two inputs connected together, but to nothing else.

Seems odd. How does that work?

User avatar
retroclinic
Posts: 3034
Joined: Thu Jul 03, 2008 1:22 pm
Location: East Riding of Yorkshire
Contact:

Re: BeebSID

Post by retroclinic » Thu Nov 12, 2009 11:09 pm

IanS wrote:IC2.2 pin 6 & IC2.3 pin 9, two inputs connected together, but to nothing else.

Seems odd. How does that work?
Well spotted, because it doesn't. They should be connected to /PGFC I beleive?

Mark.
Image

User avatar
MartinB
Posts: 5273
Joined: Mon Mar 31, 2008 9:04 pm
Location: Obscurity
Contact:

Re: BeebSID

Post by MartinB » Thu Nov 12, 2009 11:47 pm

Oh buggar, it was there until very recently :evil:

Thanks guys :D
BeebSID_s.JPG
BeebSID_s.JPG (28.29 KiB) Viewed 2794 times
I'll also do a separate 1MHz Bus named pin-list and post that tomorrow. Names on the schematic were too cluttered.

Martin

*** EDIT : Attached schematic superceded by final Issue 1 - see further down and please re-download from there ***
Attachments
BeebSID.zip
(224.03 KiB) Downloaded 94 times
Last edited by MartinB on Fri Nov 13, 2009 7:36 pm, edited 1 time in total.

User avatar
AlanD
Posts: 241
Joined: Fri Jan 09, 2009 8:30 pm
Contact:

Re: BeebSID

Post by AlanD » Thu Nov 12, 2009 11:56 pm

Helllo Martin

this fet should do the trick

Manufacturer: VISHAY FORMERLY INTERNATIONAL RECTIFIER
Farnell Order Code: 9103210
Manufacturer Part No: IRFD9014PBF

its a 4 pin DIP package so can't be mixed up with the regulators

take a look see what you think

if you look at different ones make sure the gate threshold voltage is less than about 8 to 10 volts
and look for the lowest RDS you can find
also no clamp diode on the gate

AlanD

Post Reply