Too late now, but you should have bought 100 x KSC1845FTA [ATB-ND], since the Digi-key US unit price is then only 4.7c (about 3p each!) + shippingAnyway, good luck with replacements.
Thanks for this, Ian. The postage & packing + VAT is phenomenal from the US to UK. So where these KSC's would have cost equivalent of 10p each the postage & packaging bumped it up to a unit cost of 30p. The p&p and VAT came to about £16. I only needed 20 or so - 10 (L) and 10 (R) - in the Pre Amps, but I was contemplating replacing the 2SC945's as well with KSC1845. (would that work?) so I got 80 for my £24. Wholesale Electronics Inc is HORRENDOUSLY expensive in P&P - so I wouldn't touch them with a barge pole. Michael
The 2SC945 pops up in equivalent requests all over the net because it was cheap and is a good general purpose "jelly bean" used everywhere no special requirements and only low currents are involved. It may be a GP amplifier, switch, driver, sense device, or anything. NEC says this :
The issues are:
Gain could be a bit high in some applications but since there are several different tasks here, you could try it with with some confidence. Note that 2 Hfe grades are specified again.
2SC945 is only used in the control and monitoring circuits section. There is no reason or benefit to sound in changing them so if it ain't broke, why fix it?
If you feel compelled to replace parts because you have bought replacements to do it, don't simply swap all at once in circuits that may have widely different requirements. Even simple logic says that will be the surest way to lose track of what works and what doesn't in a circuit. Where applicable, change one part in one channel at a time and test before proceeding.
The KSC1845 is not directly equivalent but will work within the ratings of 2SC945 in most applications.
The issues are:
Gain could be a bit high in some applications but since there are several different tasks here, you could try it with with some confidence. Note that 2 Hfe grades are specified again.
2SC945 is only used in the control and monitoring circuits section. There is no reason or benefit to sound in changing them so if it ain't broke, why fix it?
If you feel compelled to replace parts because you have bought replacements to do it, don't simply swap all at once in circuits that may have widely different requirements. Even simple logic says that will be the surest way to lose track of what works and what doesn't in a circuit. Where applicable, change one part in one channel at a time and test before proceeding.
Thanks, Ian. There's been a bit of a disaster here. The envelope in which the KSC1845's (all 80 of them) came was found, presumed empty and thrown out complete with contents. So I am now in a position of replacing them (again) - or - There's yet another side to this: I had earlier bought 40 X 2SC2240's as earlier comments had indicated these are a good replacement for 2SC458's and 2SC945's - so I was going to possibly replace the lot with 2SC2240's. I am currently going through them with my DCA55 and cataloguing the Hfe of each 2240 and find (I'm up to Nr.15) they are much lower than the KSC1845 ranging from 217 - 233. Do you think these 2240's would be a good replacement to the 2SC458's? There was no suffix on the Invoice, but looking at the transistor itself reads: C2240 and under that: GR 9K.
I had earlier also purchased 29 x 2SC945's again no suffix on the Invoice, but on the transistor it says: C945 and under that: Y-J11 The Hfe of one is: 159, but the legs are E, B, C. So a different order from the 2SC2240 which are E.C.B. Not having taken any of the original 945's out I don't know whether they are E,B,C or E.C.B.
Am I 'in with a chance'? Michael
You are certainly having problems there. Let's not add to them. KSC1845 is a close equivalent to the obsolete 2SC2240, which it now replaces. There isn't much between them apart from grading.
HTTP 301 This page has been moved - From the linked datasheet, gr grade is Hfe 200-400 so they are within spec. I would not try to buy more of these since being obsolete now, fakes become increasingly likely and you won't be dealing with an authorized stockist. I think they will be fine in most locations but you don't need me to repeat what has already been discussed about Hfe and whose decision this is.
As before, there is no reason to change the 2SC945 transistors, so no reason to be concerned. I don't know where you bought the replacements but as these are "no-name" products of several manufacturers, we have no clue what to expect. They could be really BC548, 2N3904 or any number of US/Euro type clones and since they are BCE pinout, I wouldn't trust any data on them since they are notorious for fakes or simply being low spec. junk - not that the OEM parts are special but you do have to be certain they will perform reliably where required, regardless of what they are and the seller's reputation.
HTTP 301 This page has been moved - From the linked datasheet, gr grade is Hfe 200-400 so they are within spec. I would not try to buy more of these since being obsolete now, fakes become increasingly likely and you won't be dealing with an authorized stockist. I think they will be fine in most locations but you don't need me to repeat what has already been discussed about Hfe and whose decision this is.
As before, there is no reason to change the 2SC945 transistors, so no reason to be concerned. I don't know where you bought the replacements but as these are "no-name" products of several manufacturers, we have no clue what to expect. They could be really BC548, 2N3904 or any number of US/Euro type clones and since they are BCE pinout, I wouldn't trust any data on them since they are notorious for fakes or simply being low spec. junk - not that the OEM parts are special but you do have to be certain they will perform reliably where required, regardless of what they are and the seller's reputation.
I have noted there are other suffixes (suffi?) to these 2SC2240's like GR 7B. There may be others. Is this an important part of the transistor number? Also whereas normally the Vbe and Ib remain constant, a few (10 out of the 40) of these 2SC2240's, the Vbe goes down from 0.76v to 0.74v. This is always accompanied by an increase in the Ib from 4.64mA to 4.66mA. In the main, the overall amplification factor of 10 2SC2240's is less than the 2SC458's that were the originals. So I am going to 'bite the bullet and re-order KSC1845's. I will first see how much P&P from Mouser is to the UK and compare it to my original order from Digi-Key. Michael
Last edited:
Thanks for the introduction to Mouser. Not happy with the DCA55 analysis of the 2SC2240's and there being overall less amplification than the original 2SC458's, I have ordered 100 KSC1845 from Mouser. Shipping charge is 30% less than Digi-Key to UK. Please don't tell me I needn't have re-ordered these KSC's! Michael
Same here
There's a seldom used feature in standard Acrobat Reader which is the "snapshot tool" which lets you frame a part of what you are seeing and turns it into a graphic file held in the Clipboard, which you can paste and edit in a Graphics editor, I use Irfanview but it canbe any other.
Now to avoid that, some PDF come as "secured" which blocks any "data extraction" , you can't even select, cut and paste text, but in that case , as suggested by Conrad, you can "print" just that page using Foxit or Cute PDF Writer into a new .pdf (give it another name or it will overwrite the original one) which by definition is "unlocked" and allows cutting parts of a schematic for graphics editing.
I do it all the time, many PDF are poorly scanned, or very low contrast, some just grey on lighter grey background, hard to see and even worse to print, so I first cut and paste them into Irfanview and then "enhance colours" by increasing contrast , rising/lowering brightness, "grayscale", "sharpen" all of which improve readability.
There's a seldom used feature in standard Acrobat Reader which is the "snapshot tool" which lets you frame a part of what you are seeing and turns it into a graphic file held in the Clipboard, which you can paste and edit in a Graphics editor, I use Irfanview but it canbe any other.
Now to avoid that, some PDF come as "secured" which blocks any "data extraction" , you can't even select, cut and paste text, but in that case , as suggested by Conrad, you can "print" just that page using Foxit or Cute PDF Writer into a new .pdf (give it another name or it will overwrite the original one) which by definition is "unlocked" and allows cutting parts of a schematic for graphics editing.
I do it all the time, many PDF are poorly scanned, or very low contrast, some just grey on lighter grey background, hard to see and even worse to print, so I first cut and paste them into Irfanview and then "enhance colours" by increasing contrast , rising/lowering brightness, "grayscale", "sharpen" all of which improve readability.
Same here
There's a seldom used feature in standard Acrobat Reader which is the "snapshot tool" which lets you frame a part of what you are seeing and turns it into a graphic file held in the Clipboard, which you can paste and edit in a Graphics editor, I use Irfanview but it canbe any other.
Now to avoid that, some PDF come as "secured" which blocks any "data extraction" , you can't even select, cut and paste text, but in that case , as suggested by Conrad, you can "print" just that page using Foxit or Cute PDF Writer into a new .pdf (give it another name or it will overwrite the original one) which by definition is "unlocked" and allows cutting parts of a schematic for graphics editing.
I do it all the time, many PDF are poorly scanned, or very low contrast, some just grey on lighter grey background, hard to see and even worse to print, so I first cut and paste them into Irfanview and then "enhance colours" by increasing contrast , rising/lowering brightness, "grayscale", "sharpen" all of which improve readability.
That's an interesting concept! When I enlarged the Cct. Diag. from the GX-635D 'Service Manual and Parts List' I found the part numbers against the components unreadable in many circumstances. Couldn't tell whether it was C17 or C14 for example. The values too were impossible to decipher. The digits were simply too blurred. The reason I was attempting this was that the parts list is incomplete. I will try your method of enhancement and thank you for the suggestion!! Michael
Some manufacturer datasheets explain their marking codes in detail, others don't. For Toshiba 2SC2240, GR is the grade which is Hfe 200-400. 7B is the batch no. or date coding so you have parts from 2 or more batches.
I have just received my box of KSC1845 transistors. All 100 of them from Mouser via FedEx. E'en as I went on-line and used FedEx 'tracking' and was told it was out for delivery the door bell rang and FedEx was at the door! Out of these 100 KSC1845 I am to choose 20 for the PreAmp Cct The original transisitors from those simple daisy-chain (L) & (R) Ccts I uploaded in Post#8 here. Since the original Cct Diag shewed the majority of transistors as 458's, there were two exceptions: TR5&TR6 which were supposed to be 945's but were in fact 458's from new, and TR9's replaced by RSBC109's from new. TR7(R) was changed to 2n2712 again from new. So someone had been at this AKAI even before it was first sold.
Question: Should I try to match all the KSC's to have the same (near-as) Hfe's as each other? No point in trying to match the old removed Hfe's is there!
So, Out with the DCA55 and start measuring and plotting 100 Hfe's!
Michael
Good to have prompt service!......So, Out with the DCA55 and start measuring and plotting 100 Hfe's!
You don't need to match these because there is nothing to match to and no requirement. The Hfe range you purchased is all the schematic calls for and that will be only a general consideration of enough without excessive gain - no need to fuss over it. For the lower Hfe types, use your 2SC2240 GRs. Simply replace the bodged parts with those specified on the schematic (or rather your replacements)
FWIW, there is no point in swapping out working parts like the 2SC945s. They are in control circuits and there is no reason they won't continue to work as new for decades to come
Thanks, Ian. I have now analysed those new KSC's. Their Hfe ranges from 352 up to 416. Since there are 100 of them I was wondering if I should look to the lower end or the higher end for the 20 required. I shall not replace any transistors other than those questionable 2SC458's. If, on completing this, it still doesn't work ... THEN I shall look at the electrolytics. Michael
Last edited:
I can see your are determined to justify using your meter here. It makes no odds if you sort to any selection within the specified range but if you are happier to have the median values then do so if you wish, as you can't go too far wrong there.
The correct approach is to use the parts as they come, without measuring and assuming the available Hfe range is the same as the spec.
The correct approach is to use the parts as they come, without measuring and assuming the available Hfe range is the same as the spec.
Just holding the transistor alters temperature and Hfe significantly - that is the truth about rubbery numbers like transistor parameters and yes, I found the Peak meters weren't that accurate, given the price (>$200 US here for the DCA75). For quicker results, a socket is best but you only discover these details after you spend your money. I now use a cheap Ebay toy that is just as good and takes no time to plug parts in and test - at around only 12 Pounds.
I would use the median values to avoid any chance of instability because the specification is actually for 2SC458BL and GR, not KSC1845F but having gone over this more than once, I leave the significance to your imagination.
I would use the median values to avoid any chance of instability because the specification is actually for 2SC458BL and GR, not KSC1845F but having gone over this more than once, I leave the significance to your imagination.
I severely lack the imagination when it comes to electronics - which is why I keep plaguing everyone (especially you, Ian, who is my very patient tutor!) about Very Basic problems I encounter. Here's one of those basic questions - about Hfe: When I removed all the old transistors Nos 1 - 10 (L) and 1 - 10 (R) I checked all their Hfe's and added them together: (L) 1-10 = 2545 gain. (R) 1 - 10 = 2691 gain. The replacement KSC1845's are, for comparison: (L) 1 - 10 = 4109 gain. (R) 1 - 10 = 4109 gain. By 'gain' I mean the total Hfe's. Well, the DCA55 did what I wanted - and I like those mini-hook 'probes', very positive connecting there - so now I have all the KSC1845's in place, guided by AKAI's Very Useful Transistor polarity markings, and the machine is just now plugged together - senza faceplate and other peripheral trimmings - ready to power on. This is the point one gets cold feet! My next post will be after the 'Power on'
fingers X'd Michael
fingers X'd Michael
Last edited:
Powered on! Every aspect of the machine functions! Checked both Line (L) & (R) IN and Mic (L) & (R) using a pair of headphones. In both the VU meters work. Now to put the casework back in place and restore to its place in the Studio over the Revox A-77. May I now express my thanks, Ian, for helping me with suggestions and very definite pointers to my happy conclusion! I am particularly grateful for the advice regarding the Electrolytics and the remainder of the Transistors. As you said "If it works why fix it!" The only recorder I have now which doesn't work is a Sony TCD-D3 - that little one. It has insurmountable problems and, due to hyper-miniaturisation, is extremely difficult to access let alone work on. So I'll stick to my TCD-D10 or the Sony PCM-R300. But I'm getting away from recording with DAT media now and have a ZOOM R16. How times have changed from my early days of EMI BTR-2! Now I have some 5" and 3" R-R tapes to Xfer to the Computer For this I will use one of my old Uher Reports - probably the 4200. Since the recordings are from ancient meetings of a railway society I doubt that in the early 60's they used stereo! . . Michael
That's very good news and suggests there wasn't too much wrong with it. I dare say though, that average circuit noise will be lower with these new transistors and that may be noticeable. Replacing the electrolytic capacitors with modern types will be the best improvement you could make to sound quality and longevity, though.
'A Good effort and I trust the next phase goes well too.
'A Good effort and I trust the next phase goes well too.
There's one extremely useful aspect to the DCA55 - since the three leads are Red, Green and Blue, it doesn't matter which colour is attached to a Transistor lead, the colour will be identified as attached to Collector, Emitter or Base. Now that's what I call useful!
Having now finished the resurrection of my AKAI GX-635D with success, I can now consider the case closed. Many thanks, Michael
Having now finished the resurrection of my AKAI GX-635D with success, I can now consider the case closed. Many thanks, Michael
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.