I got my first PowerPak in the mail today and set about testing it with my top loaders. A quick search did not reveal this to be exact duplicate info already posted, but if it is, I apologize.
Acem77figured this out on his rgb modded front loader and it applies to AV modded top loaders as well. The powerpak works fine on an unmodded nes-101, but shows only garbled graphics and sound on an AV modded top loader. The AV mod i'm using is the
darthcloudversion of the
leonkthread that is basically front loader components in place of existing top loader video out circuit. So why it doesn't work is beyond me, but after replacing the 74ls373 with a 74hct373n the powerpak works perfect.
Well the LS and HCT versions of the 74 series chips are different. Not sure what property is what fouls its use in this application.
Does one of the chips draw more current than the other or something?
I had a few strange, intermittent problems with the PowerPak on my top loader (gray blocks in the file browser, games occasionally failing to load), and it's not even modded. None of these problems are present when used with a toaster or AV Famicom. Very strange...
The 74HCT are faster than the 74LS series. That's the only difference I'm aware of.
socram8888 wrote:
The 74HCT are faster than the 74LS series. That's the only difference I'm aware of.
Not only this. LS is TTL, wich use 0..0.8v for logic 0 and 2.2..5v for logic 1. HCT is CMOS, wich use almost 0v for logic 0 and almost 5v for logic 1. Different logic technologies must be agreed between each other. For example, CPLD MAX3000 are LTTL logic (3.3v), but TTL tolerant (5.0v). You don't know, wich logic used in PowerPack, right?
PS CPU and PPU are CMOS. Other logic (address decoder, latch and RAM) are can be TTL, TTLS or CMOS.
HCT has TTL input thresholds, hence the T, and usually full-rail swinging CMOS outputs for maximum compatibility.
IIRC the PowerPak has 2.5V LVCMOS thresholds which 5V TTL outputs may or may not meet. The PowerPak definitely has pullups on all signals it must drive, but perhaps it doesn't have any on the input only address lines to assist.
socram8888 wrote:
The 74HCT are faster than the 74LS series. That's the only difference I'm aware of.
In fact it's the other way arround, LS are faster, and made of bipolar transistors, which draws much more current from the supply as opposed to MOS transistors.
There is also the difference between the '0' and '1' levels voltage range, I guess HC is more tolerant than LS but I'm not very sure.
74LS/ALS and 74HC are practically the same speed (~35 vs ~30 MHz toggle respectively). A recently made 74HC can be faster. 74F and 74AS are also similar to 74LV and slower than 74AC and 74VHC (~100 vs ~125 MHz vs ~150 MHz), so bipolar families aren't always faster apart from ECL which is historically the fastest. Plus lines are blurred with BiCMOS and there are lots of new faster MOS topologies used in modern computers.
I think TTL are more tolerant than CMOS because pure CMOS inputs must meet a high Vih (like 0.6 * VDD), but TTL inputs must only be above Vil which every logic family can meet. The issue with TTL inputs though is that you must sink a lot of current to bring the inputs low so there is a poor fanout.
HardWareMan wrote:
socram8888 wrote:
The 74HCT are faster than the 74LS series. That's the only difference I'm aware of.
Not only this. LS is TTL, wich use 0..0.8v for logic 0 and 2.2..5v for logic 1. HCT is CMOS, wich use almost 0v for logic 0 and almost 5v for logic 1.
HCT is not CMOS, but a TTL gate with CMOS-compatible inputs and outputs. It is the same as HC, but with a smaller voltage range (as it has to work with more strict inputs)
AFAIK, the LS was significantly slower than HC, and HC was slower than F (the fastest mass-produced TTL)
socram8888 wrote:
HCT is not CMOS, but a TTL gate with CMOS-compatible inputs and outputs. It is the same as HC, but with a smaller voltage range (as it has to work with more strict inputs)
Oh, really?? Say this to Texas Instruments, maybe they even don't guess. :3
So, more interesting news. The PowerPak worked fine on a different AV modded top loader, but with a Motorola 373!
The one that didn't work = HD74LS373P, not sure of manufacturer.
Motorola = SN74LS373N
HD is Hitachi and SN is Texas Instruments, not Motorola. They should be interchangeable, but the PowerPak inputs aren't biased for TTL compatibility so again it's hit or miss.
This is the 373 that the powerpak works fine with, on an AV modded nes-101.
This is the 373 that did not work with the powerpak after being AV modded.
Just found another 373 that works on an AV modded nes-101.